Compare commits

...

127 Commits

Author SHA1 Message Date
1531b6cc98 cargo: update dependencies 2024-12-13 19:42:01 +08:00
9bbc40bbfa flake: update dependencies 2024-12-13 19:41:52 +08:00
790e56d106 msys2: update 2024-12-13 19:39:39 +08:00
a00eb7969e [core] codegen: Implement matrix_power
Last of the functions that need to be ported over to strided-ndarray.
2024-12-13 15:23:31 +08:00
27a6f47330 [core] codegen: Implement construction of unsized ndarrays
Partially based on f731e604: core/ndstrides: add more ScalarOrNDArray
and NDArrayObject utils.
2024-12-13 15:23:31 +08:00
061747c67b [core] codegen: Implement NDArrayValue::atleast_nd
Based on 9cfa2622: core/ndstrides: add NDArrayObject::atleast_nd.
2024-12-13 15:23:31 +08:00
dc91d9e35a [core] codegen: Implement ScalarOrNDArray and use it in indexing
Based on 8f9d2d82: core/ndstrides: implement ndarray indexing.
2024-12-13 15:23:31 +08:00
438943ac6f [core] codegen: Implement indexing for NDArray
Based on 8f9d2d82: core/ndstrides: implement ndarray indexing

The functionality for `...` and `np.newaxis` is there in IRRT, but there
is no implementation of them for @kernel Python expressions because of
M-Labs/nac3#486.
2024-12-13 15:23:31 +08:00
678e56c95d [core] irrt: rename NDIndex to NDIndexInt
Unfortunately the name `NDIndex` is used in later commits. Renaming this
typedef to `NDIndexInt` to avoid amending. `NDIndexInt` will be removed
anyway when ndarray strides is completed.
2024-12-13 15:23:31 +08:00
fdfc80ca5f [core] codegen: Implement Slice{Type,Value}, RustSlice
Based on 01c96396: core/irrt: add Slice and Range and part of
8f9d2d82: core/ndstrides: implement ndarray indexing.

Needed for implementing general ndarray indexing.

Currently IRRT slice and range have nothing to do with NAC3's slice
and range. The IRRT slice and range are currently there to implement
ndarray specific features. However, in the future their definitions may
be used to replace that of NAC3's. (NAC3's range is a [i32 x 3], IRRT's
range is a proper struct. NAC3 does not have a slice struct).
2024-12-13 15:23:31 +08:00
8b3429d62a [artiq] Reimplement get_obj_value for strided ndarray
Based on 7ef93472: artiq: reimplement get_obj_value to use ndarray with
strides
2024-12-13 15:23:31 +08:00
f4c5038b95 [artiq] codegen: Reimplement polymorphic_print for strided ndarray
Based on 2a6ee503: artiq: reimplement polymorphic_print for ndarray
2024-12-13 15:23:31 +08:00
ddd16738a6 [core] codegen: implement ndarray iterator NDIter
Based on 50f960ab: core/ndstrides: implement ndarray iterator NDIter

A necessary utility to iterate through all elements in a possibly
strided ndarray.
2024-12-13 15:23:31 +08:00
44c49dc102 [artiq] codegen: Reimplement polymorphic_print for strided ndarray
Based on 2a6ee503: artiq: reimplement polymorphic_print for ndarray
2024-12-13 15:23:31 +08:00
e4bd376587 [core] codegen: Implement ContiguousNDArray
Fixes compatibility with linalg algorithms. matrix_power is missing due
to the need for indexing support.
2024-12-13 15:23:29 +08:00
44498f22f6 [core] codegen: Implement NDArray functions from a0a1f35b 2024-12-13 15:22:11 +08:00
110416d07a [core] codegen/irrt: Add IRRT functions for strided-ndarray 2024-12-13 15:22:11 +08:00
08a7d01a13 [core] Add itemsize and strides to NDArray struct
Temporarily disable linalg ndarray tests as they are not ported to work
with strided-ndarray.
2024-12-13 15:22:09 +08:00
3cd36fddc3 [core] codegen/types: Add check_struct_type_matches_fields
Shorthand for checking if a type is representable by a StructFields
instance.
2024-12-12 11:40:44 +08:00
56a7a9e03d [core] codegen: Add helper functions for create+call functions
Replacement for various FnCall methods from legacy ndstrides
implementation.
2024-12-12 11:30:36 +08:00
574ae40f97 [core] codegen: Add call_memcpy_generic_array
Replacement for Instance<Ptr>::copy_from from legacy ndstrides
implementation.
2024-12-12 11:30:36 +08:00
aa293b6bea [core] codegen: Add type_aligned_alloca 2024-12-12 11:30:35 +08:00
eb4b881690 [core] Expose {types,values}::ndarray modules
Allows better encapsulation of members in these modules rather than
allowing them to leak into types/values mod.
2024-12-12 11:30:14 +08:00
3d0a1d281c [core] Expose irrt::ndarray 2024-12-10 12:49:49 +08:00
ad67a99c8f [core] codegen: Cleanup builtin_fns.rs
- Unpack tuples directly in function argument
- Replace Vec parameters with slices
- Replace unwrap-transform with map-unwrap
2024-12-10 12:49:49 +08:00
8e2b50df21 [core] codegen/ndarray: Cleanup
- Remove redundant size param
- Add *_fields functions and docs
2024-12-09 13:01:08 +08:00
06092ad29b [core] Move alloca and map_value of ProxyType to implementations
These functions may not be invokable by the same set of parameters as
some classes has associated states.
2024-12-09 12:51:50 +08:00
d62c6b95fd [core] codegen/types: Rename StructField::set_from_value 2024-12-09 12:51:50 +08:00
95e29d9997 [core] codegen: Move ndarray type/value as a separate module 2024-12-09 12:51:46 +08:00
536ed2146c [meta] Remove all mentions of build_int_cast
build_int_cast performs signed extension or truncation depending on the
source and target int lengths. This is usually not what we want - We
want zero-extension instead.

Replace all instances of build_int_cast with
build_int_z_extend_or_bit_cast to fix this issue.
2024-12-09 12:51:39 +08:00
d484d44d95 [standalone] linalg: Fix function name in error message 2024-12-09 12:09:57 +08:00
ac978864f2 [meta] Apply clippy suggestions 2024-12-09 12:08:41 +08:00
95254f8464 [meta] Update Cargo dependencies 2024-12-09 12:08:41 +08:00
964945d244 string_store: update embedding map after compilation 2024-12-03 16:45:46 +08:00
ae09a0d444 exceptions: preallocate in NAC3 instead 2024-12-03 16:45:05 +08:00
01edd5af67 [meta] Apply rustfmt changes 2024-11-29 15:43:34 +08:00
015714eee1 copy constructor -> clone 2024-11-28 18:52:53 +08:00
71dec251e3 ld/dwarf: remove reader resets
DWARF reader never had to reverse. Readers are already copied to achieve this effect.
Plus the position that it reverses to might be questionable.
2024-11-28 18:52:53 +08:00
fce61f7b8c ld: fix dwarf sections offset calculations 2024-11-28 18:52:53 +08:00
babc081dbd core/toplevel: update tests 2024-11-27 14:31:57 +08:00
5337dbe23b core/toplevel: add python-like error messages for class definition 2024-11-27 14:31:57 +08:00
f862c01412 core/toplevel: refactor composer 2024-11-27 14:31:53 +08:00
0c9705f5f1 [meta] Apply clippy changes 2024-11-25 16:05:12 +08:00
5f940f86d9 [artiq] Fix obtaining ndarray struct from NDArrayType 2024-11-25 15:01:39 +08:00
5651e00688 flake: add platformdirs artiq dependency 2024-11-22 20:30:30 +08:00
f6745b987f bump sipyco and artiq used for profiling 2024-11-22 19:43:03 +08:00
e0dedc6580 nac3artiq: support kernels sent by content 2024-11-22 19:38:52 +08:00
28f574282c [core_derive] Ignore doctest in example
Causes linker errors for unknown reasons.
2024-11-22 00:00:05 +08:00
144f0922db [core] coregen/types: Implement StructFields for NDArray
Also rename some fields to better align with their naming in numpy.
2024-11-21 14:27:00 +08:00
c58ce9c3a9 [core] codegen/types: Implement NDArray in terms of i8*
Better aligns with the future implementation of ndstrides.
2024-11-21 14:27:00 +08:00
f7e296da53 [core] irrt: Break IRRT into several impl files
Each IRRT file is now mapped to one Rust file.
2024-11-21 14:27:00 +08:00
b58c99369e [core] irrt: Update some IRRT implementation
- Change CSlice to use `void*` for better pointer compatibility
- Only include impl *.hpp files in irrt.cpp
- Refactor typedef to using declaration
- Add missing ``// namespace`
2024-11-21 14:26:58 +08:00
1a535db558 [core] codegen: Add dtype to NDArrayType
We won't have this once NDArray is refactored to strided impl.
2024-11-20 15:35:57 +08:00
1ba2e287a6 [core] codegen: Add Self::llvm_type to all type abstractions 2024-11-20 15:35:57 +08:00
f95f979ad3 core/irrt: fix exception.hpp C++ castings 2024-11-20 15:35:57 +08:00
48e2148c0f core/toplevel/helper: add {extract,create}_ndims 2024-11-20 15:35:57 +08:00
88e57f7120 [core_derive] Initial implementation 2024-11-20 15:35:55 +08:00
d7633c42bc [core] codegen/types: Implement StructField{,s}
Loosely based on FieldTraversal by lyken.
2024-11-19 13:46:25 +08:00
a4f53b6e6b [core] codegen: Refactor ProxyType and ProxyValue
Accepts generator+context object for generic type checking. Also
implements more default trait impl for easier delegation.
2024-11-19 13:46:25 +08:00
9d9ead211e [core] Move Proxies to their own modules 2024-11-19 13:46:23 +08:00
26a1b85206 [core] codegen/classes: Remove Underlying type
This is confusing and we want a better abstraction than this.
2024-11-19 13:45:55 +08:00
2822074b2d [meta] Cleanup from upgrading Rust version
- Remove rust_2024_edition warnings, since it wouldn't be released for
another 3 months
- Fix new clippy warnings
2024-11-19 13:43:57 +08:00
fe67ed076c [meta] Update pre-commit configuration 2024-11-19 13:20:27 +08:00
94e2414df0 [meta] Update cargo dependencies 2024-11-19 13:20:26 +08:00
2cee760404 turn rust_2024_compatibility lints into warnings 2024-11-16 13:41:49 +08:00
230982dc84 update dependencies 2024-11-16 12:40:11 +08:00
2bd3f63991 boolop: terminate both branches with *_end_bb 2024-11-16 12:06:20 +08:00
b53266e9e6 artiq: use async RPC for attributes writeback 2024-11-12 12:04:01 +08:00
86eb22bbf3 artiq: main is always the last module 2024-11-12 12:03:38 +08:00
beaa38047d artiq: suppress main module debug warning 2024-11-12 12:03:08 +08:00
705dc4ff1c artiq: lump return value into attributes writeback RPC 2024-11-12 12:02:35 +08:00
979209a526 binop: expand not operator as loglcal not 2024-11-08 17:12:01 +08:00
c3927d0ef6 [ast] Refactor lazy_static to LazyLock
It is available in Rust 1.80 and reduces a dependency.
2024-10-30 12:29:51 +08:00
202a902cd0 [meta] Update dependencies 2024-10-30 12:29:51 +08:00
b6e2644391 [meta] Update cargo dependencies 2024-10-18 14:17:16 +08:00
45cd01556b [meta] Apply cargo fmt 2024-10-18 14:16:42 +08:00
b6cd2a6993 [meta] Reorganize order of use declarations - Phase 3 2024-10-17 16:25:52 +08:00
a98f33e6d1 [meta] Reorganize order of use declarations - Phase 2
Some more rules:

- For module-level imports, prefer no prefix > super > crate.
- Use crate instead of super if super refers to the crate-level module
2024-10-17 15:57:33 +08:00
5839badadd [standalone] Update globals.py with type-inferred global var 2024-10-07 20:44:08 +08:00
56c845aac4 [standalone] Add support for registering globals without type decl 2024-10-07 20:44:06 +08:00
65a12d9ab3 [core] Refactor registration of top-level variables 2024-10-07 17:05:48 +08:00
9c6685fa8f [core] typecheck/function_check: Fix lookup of defined ids in scope 2024-10-07 16:51:37 +08:00
2bb788e4bb [core] codegen/expr: Materialize implicit globals
Required for when globals are read without the use of a global
declaration.
2024-10-07 13:13:20 +08:00
42a2f243b5 [core] typecheck: Disallow redeclaration of var shadowing global 2024-10-07 13:11:00 +08:00
3ce2eddcdc [core] typecheck/type_inferencer: Infer whether variables are global 2024-10-07 13:10:46 +08:00
51bf126a32 [core] typecheck/type_inferencer: Differentiate global symbols
Required for analyzing use of global symbols before global declaration.
2024-10-07 12:25:00 +08:00
1a197c67f6 [core] toplevel/composer: Reduce lock scope while analyzing function 2024-10-05 15:53:20 +08:00
581b2f7bb2 [standalone] Add demo for global variables 2024-10-04 13:24:30 +08:00
746329ec5d [standalone] Implement symbol resolution for globals 2024-10-04 13:24:30 +08:00
e60e8e837f [core] Add support for global statements 2024-10-04 13:24:27 +08:00
9fdbe9695d [core] Add generator to SymbolResolver::get_symbol_value
Needed in a future commit.
2024-10-04 13:20:29 +08:00
8065e73598 [core] toplevel/composer: Add type analysis for global variables 2024-10-04 13:20:29 +08:00
192290889b [core] Add IdentifierInfo
Keeps track of whether an identifier refers to a global or local
variable.
2024-10-04 13:20:24 +08:00
1407553a2f [core] Implement parsing of global variables
Globals are now parsed into symbol resolver and top level definitions.
2024-10-04 13:18:29 +08:00
c7697606e1 [core] Add TopLevelDef::Variable 2024-10-04 13:09:25 +08:00
88d0ccbf69 [standalone] Explicit panic when encountering a compilation error
Otherwise scripts will continue to execute.
2024-10-04 13:00:16 +08:00
a43b59539c [meta] Move variables declarations closer to where they are first used 2024-10-04 13:00:16 +08:00
fe06b2806f [meta] Reorganize order of use declarations
Use declarations are now grouped into 4 groups:

- Declarations from the standard library
- Declarations from external crates
- Declarations from other crates in this project
- Declarations from within this module

Furthermore, all use declarations are grouped together to enhance
readability. super::super is also replaced by an equivalent crate::
declaration.
2024-10-04 12:52:01 +08:00
7f6c9a25ac [meta] Update Cargo dependencies 2024-10-04 12:52:01 +08:00
6c8382219f msys2: get python via numpy dependencies 2024-09-30 14:27:30 +08:00
9274a7b96b flake: update nixpkgs 2024-09-30 14:22:40 +08:00
d1c0fe2900 cargo: update dependencies 2024-09-30 14:14:43 +08:00
f2c047ba57 artiq: support async rpcs
Co-authored-by: mwojcik <mw@m-labs.hk>
Co-committed-by: mwojcik <mw@m-labs.hk>
2024-09-13 12:12:13 +08:00
5e2e77a500 [meta] Bump inkwell to v0.5 2024-09-13 11:11:14 +08:00
f3cc4702b9 [meta] Update dependencies 2024-09-13 11:11:14 +08:00
3e92c491f5 [standalone] Add tests creating ndarrays with tuple dims 2024-09-11 15:52:43 +08:00
7f629f1579 core: fix comment in unify_call 2024-09-11 15:46:19 +08:00
5640a793e2 core: allow np_full to take tuple shapes 2024-09-11 15:46:19 +08:00
abbaa506ad [standalone] Remove redundant recreation of TargetMachine 2024-09-09 14:27:10 +08:00
f3dc02d646 [meta] Apply cargo fmt 2024-09-09 14:24:52 +08:00
ea217eaea1 [meta] Update pre-commit config
Directly invoke cargo using nix develop to avoid using the system cargo.
2024-09-09 14:24:38 +08:00
5a34551905 allow the use of the LLVM shared library
Which in turns allows working around the incompatibility of the LLVM static library
with Rust link-args=-rdynamic, which produces binaries that either fail to link (OpenBSD)
or segfault on startup (Linux).

The year is 2024 and compiler toolchains are still a trash fire like this.
2024-09-09 11:17:31 +08:00
6098b1b853 fix previous commit 2024-09-06 11:32:08 +08:00
668ccb1c95 nac3core: expose inkwell and nac3parser 2024-09-06 11:06:26 +08:00
a3c624d69d update all dependencies 2024-09-06 10:21:58 +08:00
bd06155f34 irrt: compatibility with pre-C23 compilers 2024-09-05 18:54:55 +08:00
9c33c4209c [core] Fix type of ndarray.element_type
Should be the element type of the NDArray itself, not the pointer to its
type.
2024-08-30 22:47:38 +08:00
122983f11c flake: update dependencies 2024-08-30 14:45:38 +08:00
71c3a65a31 [core] codegen/stmt: Fix obtaining return type of sret functions 2024-08-29 19:15:30 +08:00
8c540d1033 [core] codegen/stmt: Add more casts for boolean types 2024-08-29 16:36:32 +08:00
0cc60a3d33 [core] codegen/expr: Fix missing cast to i1 2024-08-29 16:36:32 +08:00
a59c26aa99 [artiq] Fix RPC of ndarrays from host 2024-08-29 16:08:45 +08:00
02d93b11d1 [meta] Update dependencies 2024-08-29 14:32:21 +08:00
59cad5bfe1
standalone: clang-format demo.c 2024-08-29 10:37:24 +08:00
4318f8de84
standalone: improve src/assignment.py 2024-08-29 10:33:58 +08:00
15ac00708a [core] Use quoted include paths instead of angled brackets
This is preferred for user-defined headers.
2024-08-28 16:37:03 +08:00
c8dfdcfdea
standalone & artiq: remove class_names from resolver 2024-08-27 23:43:40 +08:00
122 changed files with 11700 additions and 5739 deletions

View File

@ -1,24 +1,24 @@
# See https://pre-commit.com for more information
# See https://pre-commit.com/hooks.html for more hooks
default_stages: [commit]
default_stages: [pre-commit]
repos:
- repo: local
hooks:
- id: nac3-cargo-fmt
name: nac3 cargo format
entry: cargo
entry: nix
language: system
types: [file, rust]
pass_filenames: false
description: Runs cargo fmt on the codebase.
args: [fmt]
args: [develop, -c, cargo, fmt, --all]
- id: nac3-cargo-clippy
name: nac3 cargo clippy
entry: cargo
entry: nix
language: system
types: [file, rust]
pass_filenames: false
description: Runs cargo clippy on the codebase.
args: [clippy, --tests]
args: [develop, -c, cargo, clippy, --tests]

529
Cargo.lock generated

File diff suppressed because it is too large Load Diff

View File

@ -4,6 +4,7 @@ members = [
"nac3ast",
"nac3parser",
"nac3core",
"nac3core/nac3core_derive",
"nac3standalone",
"nac3artiq",
"runkernel",

6
flake.lock generated
View File

@ -2,11 +2,11 @@
"nodes": {
"nixpkgs": {
"locked": {
"lastModified": 1723637854,
"narHash": "sha256-med8+5DSWa2UnOqtdICndjDAEjxr5D7zaIiK4pn0Q7c=",
"lastModified": 1733940404,
"narHash": "sha256-Pj39hSoUA86ZePPF/UXiYHHM7hMIkios8TYG29kQT4g=",
"owner": "NixOS",
"repo": "nixpkgs",
"rev": "c3aa7b8938b17aebd2deecf7be0636000d62a2b9",
"rev": "5d67ea6b4b63378b9c13be21e2ec9d1afc921713",
"type": "github"
},
"original": {

View File

@ -107,18 +107,18 @@
(pkgs.fetchFromGitHub {
owner = "m-labs";
repo = "sipyco";
rev = "939f84f9b5eef7efbf7423c735d1834783b6140e";
sha256 = "sha256-15Nun4EY35j+6SPZkjzZtyH/ncxLS60KuGJjFh5kSTc=";
rev = "094a6cd63ffa980ef63698920170e50dc9ba77fd";
sha256 = "sha256-PPnAyDedUQ7Og/Cby9x5OT9wMkNGTP8GS53V6N/dk4w=";
})
(pkgs.fetchFromGitHub {
owner = "m-labs";
repo = "artiq";
rev = "923ca3377d42c815f979983134ec549dc39d3ca0";
sha256 = "sha256-oJoEeNEeNFSUyh6jXG8Tzp6qHVikeHS0CzfE+mODPgw=";
rev = "28c9de3e251daa89a8c9fd79d5ab64a3ec03bac6";
sha256 = "sha256-vAvpbHc5B+1wtG8zqN7j9dQE1ON+i22v+uqA+tw6Gak=";
})
];
buildInputs = [
(python3-mimalloc.withPackages(ps: [ ps.numpy ps.scipy ps.jsonschema ps.lmdb nac3artiq-instrumented ]))
(python3-mimalloc.withPackages(ps: [ ps.numpy ps.scipy ps.jsonschema ps.lmdb ps.platformdirs nac3artiq-instrumented ]))
pkgs.llvmPackages_14.llvm.out
];
phases = [ "buildPhase" "installPhase" ];

View File

@ -12,16 +12,10 @@ crate-type = ["cdylib"]
itertools = "0.13"
pyo3 = { version = "0.21", features = ["extension-module", "gil-refs"] }
parking_lot = "0.12"
tempfile = "3.10"
nac3parser = { path = "../nac3parser" }
tempfile = "3.13"
nac3core = { path = "../nac3core" }
nac3ld = { path = "../nac3ld" }
[dependencies.inkwell]
version = "0.4"
default-features = false
features = ["llvm14-0", "target-x86", "target-arm", "target-riscv", "no-libffi-linking"]
[features]
init-llvm-profile = []
no-escape-analysis = ["nac3core/no-escape-analysis"]

View File

@ -7,33 +7,6 @@ class EmbeddingMap:
self.function_map = {}
self.attributes_writeback = []
# preallocate exception names
self.preallocate_runtime_exception_names(["RuntimeError",
"RTIOUnderflow",
"RTIOOverflow",
"RTIODestinationUnreachable",
"DMAError",
"I2CError",
"CacheError",
"SPIError",
"0:ZeroDivisionError",
"0:IndexError",
"0:ValueError",
"0:RuntimeError",
"0:AssertionError",
"0:KeyError",
"0:NotImplementedError",
"0:OverflowError",
"0:IOError",
"0:UnwrapNoneError"])
def preallocate_runtime_exception_names(self, names):
for i, name in enumerate(names):
if ":" not in name:
name = "0:artiq.coredevice.exceptions." + name
exn_id = self.store_str(name)
assert exn_id == i
def store_function(self, key, fun):
self.function_map[key] = fun
return key

View File

@ -112,10 +112,15 @@ def extern(function):
register_function(function)
return function
def rpc(function):
"""Decorates a function declaration defined by the core device runtime."""
register_function(function)
return function
def rpc(arg=None, flags={}):
"""Decorates a function or method to be executed on the host interpreter."""
if arg is None:
def inner_decorator(function):
return rpc(function, flags)
return inner_decorator
register_function(arg)
return arg
def kernel(function_or_method):
"""Decorates a function or method to be executed on the core device."""
@ -201,7 +206,7 @@ class Core:
embedding = EmbeddingMap()
if allow_registration:
compiler.analyze(registered_functions, registered_classes)
compiler.analyze(registered_functions, registered_classes, set())
allow_registration = False
if hasattr(method, "__self__"):

View File

@ -1,38 +1,3 @@
use nac3core::{
codegen::{
classes::{
ArrayLikeIndexer, ArrayLikeValue, ArraySliceValue, ListValue, NDArrayType,
NDArrayValue, RangeValue, UntypedArrayLikeAccessor,
},
expr::{destructure_range, gen_call},
irrt::call_ndarray_calc_size,
llvm_intrinsics::{call_int_smax, call_memcpy_generic, call_stackrestore, call_stacksave},
stmt::{gen_block, gen_for_callback_incrementing, gen_if_callback, gen_with},
CodeGenContext, CodeGenerator,
},
symbol_resolver::ValueEnum,
toplevel::{helper::PrimDef, numpy::unpack_ndarray_var_tys, DefinitionId, GenCall},
typecheck::typedef::{iter_type_vars, FunSignature, FuncArg, Type, TypeEnum, VarMap},
};
use nac3parser::ast::{Expr, ExprKind, Located, Stmt, StmtKind, StrRef};
use inkwell::{
context::Context,
module::Linkage,
types::{BasicType, IntType},
values::{BasicValueEnum, PointerValue, StructValue},
AddressSpace, IntPredicate,
};
use pyo3::{
types::{PyDict, PyList},
PyObject, PyResult, Python,
};
use crate::{symbol_resolver::InnerResolver, timeline::TimeFns};
use itertools::Itertools;
use std::{
collections::{hash_map::DefaultHasher, HashMap},
hash::{Hash, Hasher},
@ -41,6 +6,43 @@ use std::{
sync::Arc,
};
use itertools::Itertools;
use pyo3::{
types::{PyDict, PyList},
PyObject, PyResult, Python,
};
use super::{symbol_resolver::InnerResolver, timeline::TimeFns};
use nac3core::{
codegen::{
expr::{destructure_range, gen_call},
llvm_intrinsics::{call_int_smax, call_memcpy, call_stackrestore, call_stacksave},
stmt::{gen_block, gen_for_callback_incrementing, gen_if_callback, gen_with},
type_aligned_alloca,
types::ndarray::NDArrayType,
values::{
ArrayLikeIndexer, ArrayLikeValue, ArraySliceValue, ListValue, ProxyValue, RangeValue,
UntypedArrayLikeAccessor,
},
CodeGenContext, CodeGenerator,
},
inkwell::{
context::Context,
module::Linkage,
types::{BasicType, IntType},
values::{BasicValueEnum, IntValue, PointerValue, StructValue},
AddressSpace, IntPredicate, OptimizationLevel,
},
nac3parser::ast::{Expr, ExprKind, Located, Stmt, StmtKind, StrRef},
symbol_resolver::ValueEnum,
toplevel::{
helper::{extract_ndims, PrimDef},
numpy::unpack_ndarray_var_tys,
DefinitionId, GenCall,
},
typecheck::typedef::{iter_type_vars, FunSignature, FuncArg, Type, TypeEnum, VarMap},
};
/// The parallelism mode within a block.
#[derive(Copy, Clone, Eq, PartialEq)]
enum ParallelMode {
@ -459,55 +461,49 @@ fn format_rpc_arg<'ctx>(
let llvm_i1 = ctx.ctx.bool_type();
let llvm_usize = generator.get_size_type(ctx.ctx);
let (elem_ty, _) = unpack_ndarray_var_tys(&mut ctx.unifier, arg_ty);
let llvm_arg_ty =
NDArrayType::new(generator, ctx.ctx, ctx.get_llvm_type(generator, elem_ty));
let llvm_arg = NDArrayValue::from_ptr_val(arg.into_pointer_value(), llvm_usize, None);
let (elem_ty, ndims) = unpack_ndarray_var_tys(&mut ctx.unifier, arg_ty);
let ndims = extract_ndims(&ctx.unifier, ndims);
let dtype = ctx.get_llvm_type(generator, elem_ty);
let ndarray = NDArrayType::new(generator, ctx.ctx, dtype, Some(ndims))
.map_value(arg.into_pointer_value(), None);
let llvm_usize_sizeof = ctx
.builder
.build_int_truncate_or_bit_cast(llvm_arg_ty.size_type().size_of(), llvm_usize, "")
.unwrap();
let llvm_pdata_sizeof = ctx
.builder
.build_int_truncate_or_bit_cast(
llvm_arg_ty.element_type().ptr_type(AddressSpace::default()).size_of(),
llvm_usize,
"",
)
.unwrap();
let ndims = llvm_usize.const_int(ndims, false);
let dims_buf_sz =
ctx.builder.build_int_mul(llvm_arg.load_ndims(ctx), llvm_usize_sizeof, "").unwrap();
// `ndarray.data` is possibly not contiguous, and we need it to be contiguous for
// the reader.
// Turning it into a ContiguousNDArray to get a `data` that is contiguous.
let carray = ndarray.make_contiguous_ndarray(generator, ctx);
let buffer_size =
ctx.builder.build_int_add(dims_buf_sz, llvm_pdata_sizeof, "").unwrap();
let sizeof_usize = llvm_usize.size_of();
let sizeof_usize =
ctx.builder.build_int_z_extend_or_bit_cast(sizeof_usize, llvm_usize, "").unwrap();
let buffer = ctx.builder.build_array_alloca(llvm_i8, buffer_size, "rpc.arg").unwrap();
let buffer = ArraySliceValue::from_ptr_val(buffer, buffer_size, Some("rpc.arg"));
let sizeof_pdata = dtype.ptr_type(AddressSpace::default()).size_of();
let sizeof_pdata =
ctx.builder.build_int_z_extend_or_bit_cast(sizeof_pdata, llvm_usize, "").unwrap();
let ppdata = generator.gen_var_alloc(ctx, llvm_arg_ty.element_type(), None).unwrap();
ctx.builder.build_store(ppdata, llvm_arg.data().base_ptr(ctx, generator)).unwrap();
let sizeof_buf_shape = ctx.builder.build_int_mul(sizeof_usize, ndims, "").unwrap();
let sizeof_buf = ctx.builder.build_int_add(sizeof_buf_shape, sizeof_pdata, "").unwrap();
call_memcpy_generic(
ctx,
buffer.base_ptr(ctx, generator),
ppdata,
llvm_pdata_sizeof,
llvm_i1.const_zero(),
);
// buf = { data: void*, shape: [size_t; ndims]; }
let buf = ctx.builder.build_array_alloca(llvm_i8, sizeof_buf, "rpc.arg").unwrap();
let buf = ArraySliceValue::from_ptr_val(buf, sizeof_buf, Some("rpc.arg"));
let buf_data = buf.base_ptr(ctx, generator);
let buf_shape =
unsafe { buf.ptr_offset_unchecked(ctx, generator, &sizeof_pdata, None) };
let pbuffer_dims_begin =
unsafe { buffer.ptr_offset_unchecked(ctx, generator, &llvm_pdata_sizeof, None) };
call_memcpy_generic(
ctx,
pbuffer_dims_begin,
llvm_arg.dim_sizes().base_ptr(ctx, generator),
dims_buf_sz,
llvm_i1.const_zero(),
);
// Write to `buf->data`
let carray_data = carray.load_data(ctx);
let carray_data = ctx.builder.build_pointer_cast(carray_data, llvm_pi8, "").unwrap();
call_memcpy(ctx, buf_data, carray_data, sizeof_pdata, llvm_i1.const_zero());
buffer.base_ptr(ctx, generator)
// Write to `buf->shape`
let carray_shape = ndarray.shape().base_ptr(ctx, generator);
let carray_shape_i8 =
ctx.builder.build_pointer_cast(carray_shape, llvm_pi8, "").unwrap();
call_memcpy(ctx, buf_shape, carray_shape_i8, sizeof_buf_shape, llvm_i1.const_zero());
buf.base_ptr(ctx, generator)
}
_ => {
@ -517,7 +513,7 @@ fn format_rpc_arg<'ctx>(
ctx.builder.build_store(arg_slot, arg).unwrap();
ctx.builder
.build_bitcast(arg_slot, llvm_pi8, "rpc.arg")
.build_bit_cast(arg_slot, llvm_pi8, "rpc.arg")
.map(BasicValueEnum::into_pointer_value)
.unwrap()
}
@ -528,12 +524,276 @@ fn format_rpc_arg<'ctx>(
arg_slot
}
/// Formats an RPC return value to conform to the expected format required by NAC3.
fn format_rpc_ret<'ctx>(
generator: &mut dyn CodeGenerator,
ctx: &mut CodeGenContext<'ctx, '_>,
ret_ty: Type,
) -> Option<BasicValueEnum<'ctx>> {
// -- receive value:
// T result = {
// void *ret_ptr = alloca(sizeof(T));
// void *ptr = ret_ptr;
// loop: int size = rpc_recv(ptr);
// // Non-zero: Provide `size` bytes of extra storage for variable-length data.
// if(size) { ptr = alloca(size); goto loop; }
// else *(T*)ret_ptr
// }
let llvm_i8 = ctx.ctx.i8_type();
let llvm_i32 = ctx.ctx.i32_type();
let llvm_i8_8 = ctx.ctx.struct_type(&[llvm_i8.array_type(8).into()], false);
let llvm_pi8 = llvm_i8.ptr_type(AddressSpace::default());
let llvm_usize = generator.get_size_type(ctx.ctx);
let llvm_pusize = llvm_usize.ptr_type(AddressSpace::default());
let rpc_recv = ctx.module.get_function("rpc_recv").unwrap_or_else(|| {
ctx.module.add_function("rpc_recv", llvm_i32.fn_type(&[llvm_pi8.into()], false), None)
});
if ctx.unifier.unioned(ret_ty, ctx.primitives.none) {
ctx.build_call_or_invoke(rpc_recv, &[llvm_pi8.const_null().into()], "rpc_recv");
return None;
}
let prehead_bb = ctx.builder.get_insert_block().unwrap();
let current_function = prehead_bb.get_parent().unwrap();
let head_bb = ctx.ctx.append_basic_block(current_function, "rpc.head");
let alloc_bb = ctx.ctx.append_basic_block(current_function, "rpc.continue");
let tail_bb = ctx.ctx.append_basic_block(current_function, "rpc.tail");
let llvm_ret_ty = ctx.get_llvm_abi_type(generator, ret_ty);
let result = match &*ctx.unifier.get_ty_immutable(ret_ty) {
TypeEnum::TObj { obj_id, .. } if *obj_id == PrimDef::NDArray.id() => {
let num_0 = llvm_usize.const_zero();
// Round `val` up to its modulo `power_of_two`
let round_up = |ctx: &mut CodeGenContext<'ctx, '_>,
val: IntValue<'ctx>,
power_of_two: IntValue<'ctx>| {
debug_assert_eq!(
val.get_type().get_bit_width(),
power_of_two.get_type().get_bit_width()
);
let llvm_val_t = val.get_type();
let max_rem = ctx
.builder
.build_int_sub(power_of_two, llvm_val_t.const_int(1, false), "")
.unwrap();
ctx.builder
.build_and(
ctx.builder.build_int_add(val, max_rem, "").unwrap(),
ctx.builder.build_not(max_rem, "").unwrap(),
"",
)
.unwrap()
};
// Allocate the resulting ndarray
// A condition after format_rpc_ret ensures this will not be popped this off.
let (dtype, ndims) = unpack_ndarray_var_tys(&mut ctx.unifier, ret_ty);
let dtype_llvm = ctx.get_llvm_type(generator, dtype);
let ndims = extract_ndims(&ctx.unifier, ndims);
let ndarray = NDArrayType::new(generator, ctx.ctx, dtype_llvm, Some(ndims))
.construct_uninitialized(generator, ctx, None);
// NOTE: Current content of `ndarray`:
// - * `data` - **NOT YET** allocated.
// - * `itemsize` - initialized to be size_of(dtype).
// - * `ndims` - initialized.
// - * `shape` - allocated; has uninitialized values.
// - * `strides` - allocated; has uninitialized values.
let itemsize = ndarray.load_itemsize(ctx); // Same as doing a `ctx.get_llvm_type` on `dtype` and get its `size_of()`.
// Allocates a buffer for the initial RPC'ed object, which is guaranteed to be
// (4 + 4 * ndims) bytes with 8-byte alignment
let sizeof_usize = llvm_usize.size_of();
let sizeof_usize =
ctx.builder.build_int_truncate_or_bit_cast(sizeof_usize, llvm_usize, "").unwrap();
let sizeof_ptr = llvm_i8.ptr_type(AddressSpace::default()).size_of();
let sizeof_ptr =
ctx.builder.build_int_z_extend_or_bit_cast(sizeof_ptr, llvm_usize, "").unwrap();
let sizeof_shape =
ctx.builder.build_int_mul(ndarray.load_ndims(ctx), sizeof_usize, "").unwrap();
// Size of the buffer for the initial `rpc_recv()`.
let unaligned_buffer_size =
ctx.builder.build_int_add(sizeof_ptr, sizeof_shape, "").unwrap();
let stackptr = call_stacksave(ctx, None);
let buffer = type_aligned_alloca(
generator,
ctx,
llvm_i8_8,
unaligned_buffer_size,
Some("rpc.buffer"),
);
let buffer = ArraySliceValue::from_ptr_val(buffer, unaligned_buffer_size, None);
// The first call to `rpc_recv` reads the top-level ndarray object: [pdata, shape]
//
// The returned value is the number of bytes for `ndarray.data`.
let ndarray_nbytes = ctx
.build_call_or_invoke(
rpc_recv,
&[buffer.base_ptr(ctx, generator).into()], // Reads [usize; ndims]
"rpc.size.next",
)
.map(BasicValueEnum::into_int_value)
.unwrap();
// debug_assert(ndarray_nbytes > 0)
if ctx.registry.llvm_options.opt_level == OptimizationLevel::None {
let cmp = ctx
.builder
.build_int_compare(IntPredicate::UGT, ndarray_nbytes, num_0, "")
.unwrap();
ctx.make_assert(
generator,
cmp,
"0:AssertionError",
"Unexpected RPC termination for ndarray - Expected data buffer next",
[None, None, None],
ctx.current_loc,
);
}
// Copy shape from the buffer to `ndarray.shape`.
// We need to skip the first `sizeof(uint8_t*)` bytes to skip the `pdata` in `[pdata, shape]`.
let pbuffer_shape =
unsafe { buffer.ptr_offset_unchecked(ctx, generator, &sizeof_ptr, None) };
let pbuffer_shape =
ctx.builder.build_pointer_cast(pbuffer_shape, llvm_pusize, "").unwrap();
// Copy shape from buffer to `ndarray.shape`
ndarray.copy_shape_from_array(generator, ctx, pbuffer_shape);
// Restore stack from before allocation of buffer
call_stackrestore(ctx, stackptr);
// Allocate `ndarray.data`.
// `ndarray.shape` must be initialized beforehand in this implementation
// (for ndarray.create_data() to know how many elements to allocate)
unsafe { ndarray.create_data(generator, ctx) }; // NOTE: the strides of `ndarray` has also been set to contiguous in `create_data`.
// debug_assert(nelems * sizeof(T) >= ndarray_nbytes)
if ctx.registry.llvm_options.opt_level == OptimizationLevel::None {
let num_elements = ndarray.size(generator, ctx);
let expected_ndarray_nbytes =
ctx.builder.build_int_mul(num_elements, itemsize, "").unwrap();
let cmp = ctx
.builder
.build_int_compare(
IntPredicate::UGE,
expected_ndarray_nbytes,
ndarray_nbytes,
"",
)
.unwrap();
ctx.make_assert(
generator,
cmp,
"0:AssertionError",
"Unexpected allocation size request for ndarray data - Expected up to {0} bytes, got {1} bytes",
[Some(expected_ndarray_nbytes), Some(ndarray_nbytes), None],
ctx.current_loc,
);
}
let ndarray_data = ndarray.data().base_ptr(ctx, generator);
// NOTE: Currently on `prehead_bb`
ctx.builder.build_unconditional_branch(head_bb).unwrap();
// Inserting into `head_bb`. Do `rpc_recv` for `data` recursively.
ctx.builder.position_at_end(head_bb);
let phi = ctx.builder.build_phi(llvm_pi8, "rpc.ptr").unwrap();
phi.add_incoming(&[(&ndarray_data, prehead_bb)]);
let alloc_size = ctx
.build_call_or_invoke(rpc_recv, &[phi.as_basic_value()], "rpc.size.next")
.map(BasicValueEnum::into_int_value)
.unwrap();
let is_done = ctx
.builder
.build_int_compare(IntPredicate::EQ, llvm_i32.const_zero(), alloc_size, "rpc.done")
.unwrap();
ctx.builder.build_conditional_branch(is_done, tail_bb, alloc_bb).unwrap();
ctx.builder.position_at_end(alloc_bb);
// Align the allocation to sizeof(T)
let alloc_size = round_up(ctx, alloc_size, itemsize);
// TODO(Derppening): Candidate for refactor into type_aligned_alloca
let alloc_ptr = ctx
.builder
.build_array_alloca(
dtype_llvm,
ctx.builder.build_int_unsigned_div(alloc_size, itemsize, "").unwrap(),
"rpc.alloc",
)
.unwrap();
let alloc_ptr =
ctx.builder.build_pointer_cast(alloc_ptr, llvm_pi8, "rpc.alloc.ptr").unwrap();
phi.add_incoming(&[(&alloc_ptr, alloc_bb)]);
ctx.builder.build_unconditional_branch(head_bb).unwrap();
ctx.builder.position_at_end(tail_bb);
ndarray.as_base_value().into()
}
_ => {
let slot = ctx.builder.build_alloca(llvm_ret_ty, "rpc.ret.slot").unwrap();
let slotgen = ctx.builder.build_bit_cast(slot, llvm_pi8, "rpc.ret.ptr").unwrap();
ctx.builder.build_unconditional_branch(head_bb).unwrap();
ctx.builder.position_at_end(head_bb);
let phi = ctx.builder.build_phi(llvm_pi8, "rpc.ptr").unwrap();
phi.add_incoming(&[(&slotgen, prehead_bb)]);
let alloc_size = ctx
.build_call_or_invoke(rpc_recv, &[phi.as_basic_value()], "rpc.size.next")
.unwrap()
.into_int_value();
let is_done = ctx
.builder
.build_int_compare(IntPredicate::EQ, llvm_i32.const_zero(), alloc_size, "rpc.done")
.unwrap();
ctx.builder.build_conditional_branch(is_done, tail_bb, alloc_bb).unwrap();
ctx.builder.position_at_end(alloc_bb);
let alloc_ptr =
ctx.builder.build_array_alloca(llvm_pi8, alloc_size, "rpc.alloc").unwrap();
let alloc_ptr =
ctx.builder.build_bit_cast(alloc_ptr, llvm_pi8, "rpc.alloc.ptr").unwrap();
phi.add_incoming(&[(&alloc_ptr, alloc_bb)]);
ctx.builder.build_unconditional_branch(head_bb).unwrap();
ctx.builder.position_at_end(tail_bb);
ctx.builder.build_load(slot, "rpc.result").unwrap()
}
};
Some(result)
}
fn rpc_codegen_callback_fn<'ctx>(
ctx: &mut CodeGenContext<'ctx, '_>,
obj: Option<(Type, ValueEnum<'ctx>)>,
fun: (&FunSignature, DefinitionId),
args: Vec<(Option<StrRef>, ValueEnum<'ctx>)>,
generator: &mut dyn CodeGenerator,
is_async: bool,
) -> Result<Option<BasicValueEnum<'ctx>>, String> {
let int8 = ctx.ctx.i8_type();
let int32 = ctx.ctx.i32_type();
@ -642,6 +902,29 @@ fn rpc_codegen_callback_fn<'ctx>(
}
// call
if is_async {
let rpc_send_async = ctx.module.get_function("rpc_send_async").unwrap_or_else(|| {
ctx.module.add_function(
"rpc_send_async",
ctx.ctx.void_type().fn_type(
&[
int32.into(),
tag_ptr_type.ptr_type(AddressSpace::default()).into(),
ptr_type.ptr_type(AddressSpace::default()).into(),
],
false,
),
None,
)
});
ctx.builder
.build_call(
rpc_send_async,
&[service_id.into(), tag_ptr.into(), args_ptr.into()],
"rpc.send",
)
.unwrap();
} else {
let rpc_send = ctx.module.get_function("rpc_send").unwrap_or_else(|| {
ctx.module.add_function(
"rpc_send",
@ -659,74 +942,32 @@ fn rpc_codegen_callback_fn<'ctx>(
ctx.builder
.build_call(rpc_send, &[service_id.into(), tag_ptr.into(), args_ptr.into()], "rpc.send")
.unwrap();
}
// reclaim stack space used by arguments
call_stackrestore(ctx, stackptr);
// -- receive value:
// T result = {
// void *ret_ptr = alloca(sizeof(T));
// void *ptr = ret_ptr;
// loop: int size = rpc_recv(ptr);
// // Non-zero: Provide `size` bytes of extra storage for variable-length data.
// if(size) { ptr = alloca(size); goto loop; }
// else *(T*)ret_ptr
// }
let rpc_recv = ctx.module.get_function("rpc_recv").unwrap_or_else(|| {
ctx.module.add_function("rpc_recv", int32.fn_type(&[ptr_type.into()], false), None)
});
if is_async {
// async RPCs do not return any values
Ok(None)
} else {
let result = format_rpc_ret(generator, ctx, fun.0.ret);
if ctx.unifier.unioned(fun.0.ret, ctx.primitives.none) {
ctx.build_call_or_invoke(rpc_recv, &[ptr_type.const_null().into()], "rpc_recv");
return Ok(None);
}
let prehead_bb = ctx.builder.get_insert_block().unwrap();
let current_function = prehead_bb.get_parent().unwrap();
let head_bb = ctx.ctx.append_basic_block(current_function, "rpc.head");
let alloc_bb = ctx.ctx.append_basic_block(current_function, "rpc.continue");
let tail_bb = ctx.ctx.append_basic_block(current_function, "rpc.tail");
let ret_ty = ctx.get_llvm_abi_type(generator, fun.0.ret);
let need_load = !ret_ty.is_pointer_type();
let slot = ctx.builder.build_alloca(ret_ty, "rpc.ret.slot").unwrap();
let slotgen = ctx.builder.build_bitcast(slot, ptr_type, "rpc.ret.ptr").unwrap();
ctx.builder.build_unconditional_branch(head_bb).unwrap();
ctx.builder.position_at_end(head_bb);
let phi = ctx.builder.build_phi(ptr_type, "rpc.ptr").unwrap();
phi.add_incoming(&[(&slotgen, prehead_bb)]);
let alloc_size = ctx
.build_call_or_invoke(rpc_recv, &[phi.as_basic_value()], "rpc.size.next")
.unwrap()
.into_int_value();
let is_done = ctx
.builder
.build_int_compare(inkwell::IntPredicate::EQ, int32.const_zero(), alloc_size, "rpc.done")
.unwrap();
ctx.builder.build_conditional_branch(is_done, tail_bb, alloc_bb).unwrap();
ctx.builder.position_at_end(alloc_bb);
let alloc_ptr = ctx.builder.build_array_alloca(ptr_type, alloc_size, "rpc.alloc").unwrap();
let alloc_ptr = ctx.builder.build_bitcast(alloc_ptr, ptr_type, "rpc.alloc.ptr").unwrap();
phi.add_incoming(&[(&alloc_ptr, alloc_bb)]);
ctx.builder.build_unconditional_branch(head_bb).unwrap();
ctx.builder.position_at_end(tail_bb);
let result = ctx.builder.build_load(slot, "rpc.result").unwrap();
if need_load {
if !result.is_some_and(|res| res.get_type().is_pointer_type()) {
// An RPC returning an NDArray would not touch here.
call_stackrestore(ctx, stackptr);
}
Ok(Some(result))
Ok(result)
}
}
pub fn attributes_writeback(
ctx: &mut CodeGenContext<'_, '_>,
pub fn attributes_writeback<'ctx>(
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut dyn CodeGenerator,
inner_resolver: &InnerResolver,
host_attributes: &PyObject,
return_obj: Option<(Type, ValueEnum<'ctx>)>,
) -> Result<(), String> {
Python::with_gil(|py| -> PyResult<Result<(), String>> {
let host_attributes: &PyList = host_attributes.downcast(py)?;
@ -736,6 +977,11 @@ pub fn attributes_writeback(
let zero = int32.const_zero();
let mut values = Vec::new();
let mut scratch_buffer = Vec::new();
if let Some((ty, obj)) = return_obj {
values.push((ty, obj.to_basic_value_enum(ctx, generator, ty).unwrap()));
}
for val in (*globals).values() {
let val = val.as_ref(py);
let ty = inner_resolver.get_obj_type(
@ -814,7 +1060,7 @@ pub fn attributes_writeback(
let args: Vec<_> =
values.into_iter().map(|(_, val)| (None, ValueEnum::Dynamic(val))).collect();
if let Err(e) =
rpc_codegen_callback_fn(ctx, None, (&fun, PrimDef::Int32.id()), args, generator)
rpc_codegen_callback_fn(ctx, None, (&fun, PrimDef::Int32.id()), args, generator, true)
{
return Ok(Err(e));
}
@ -824,9 +1070,9 @@ pub fn attributes_writeback(
Ok(())
}
pub fn rpc_codegen_callback() -> Arc<GenCall> {
Arc::new(GenCall::new(Box::new(|ctx, obj, fun, args, generator| {
rpc_codegen_callback_fn(ctx, obj, fun, args, generator)
pub fn rpc_codegen_callback(is_async: bool) -> Arc<GenCall> {
Arc::new(GenCall::new(Box::new(move |ctx, obj, fun, args, generator| {
rpc_codegen_callback_fn(ctx, obj, fun, args, generator, is_async)
})))
}
@ -1040,7 +1286,8 @@ fn polymorphic_print<'ctx>(
fmt.push('[');
flush(ctx, generator, &mut fmt, &mut args);
let val = ListValue::from_ptr_val(value.into_pointer_value(), llvm_usize, None);
let val =
ListValue::from_pointer_value(value.into_pointer_value(), llvm_usize, None);
let len = val.load_size(ctx, None);
let last =
ctx.builder.build_int_sub(len, llvm_usize.const_int(1, false), "").unwrap();
@ -1091,56 +1338,50 @@ fn polymorphic_print<'ctx>(
}
TypeEnum::TObj { obj_id, .. } if *obj_id == PrimDef::NDArray.id() => {
let (elem_ty, _) = unpack_ndarray_var_tys(&mut ctx.unifier, ty);
fmt.push_str("array([");
flush(ctx, generator, &mut fmt, &mut args);
let val = NDArrayValue::from_ptr_val(value.into_pointer_value(), llvm_usize, None);
let len = call_ndarray_calc_size(generator, ctx, &val.dim_sizes(), (None, None));
let last =
ctx.builder.build_int_sub(len, llvm_usize.const_int(1, false), "").unwrap();
let (dtype, _) = unpack_ndarray_var_tys(&mut ctx.unifier, ty);
let ndarray = NDArrayType::from_unifier_type(generator, ctx, ty)
.map_value(value.into_pointer_value(), None);
gen_for_callback_incrementing(
generator,
ctx,
None,
llvm_usize.const_zero(),
(len, false),
|generator, ctx, _, i| {
let elem = unsafe { val.data().get_unchecked(ctx, generator, &i, None) };
let num_0 = llvm_usize.const_zero();
polymorphic_print(
ctx,
generator,
&[(elem_ty, elem.into())],
"",
None,
true,
as_rtio,
)?;
// Print `ndarray` as a flat list delimited by interspersed with ", \0"
ndarray.foreach(generator, ctx, |generator, ctx, _, hdl| {
let i = hdl.get_index(ctx);
let scalar = hdl.get_scalar(ctx);
// if (i != 0) puts(", ");
gen_if_callback(
generator,
ctx,
|_, ctx| {
Ok(ctx
let not_first = ctx
.builder
.build_int_compare(IntPredicate::ULT, i, last, "")
.unwrap())
.build_int_compare(IntPredicate::NE, i, num_0, "")
.unwrap();
Ok(not_first)
},
|generator, ctx| {
printf(ctx, generator, ", \0".into(), Vec::default());
Ok(())
},
|_, _| Ok(()),
)?;
Ok(())
},
llvm_usize.const_int(1, false),
// Print element
polymorphic_print(
ctx,
generator,
&[(dtype, scalar.into())],
"",
None,
true,
as_rtio,
)?;
Ok(())
})?;
fmt.push_str(")]");
flush(ctx, generator, &mut fmt, &mut args);
@ -1150,7 +1391,7 @@ fn polymorphic_print<'ctx>(
fmt.push_str("range(");
flush(ctx, generator, &mut fmt, &mut args);
let val = RangeValue::from_ptr_val(value.into_pointer_value(), None);
let val = RangeValue::from_pointer_value(value.into_pointer_value(), None);
let (start, stop, step) = destructure_range(ctx, val);

View File

@ -1,10 +1,4 @@
#![deny(
future_incompatible,
let_underscore,
nonstandard_style,
rust_2024_compatibility,
clippy::all
)]
#![deny(future_incompatible, let_underscore, nonstandard_style, clippy::all)]
#![warn(clippy::pedantic)]
#![allow(
unsafe_op_in_unsafe_fn,
@ -16,64 +10,65 @@
clippy::wildcard_imports
)]
use std::collections::{HashMap, HashSet};
use std::fs;
use std::io::Write;
use std::process::Command;
use std::rc::Rc;
use std::sync::Arc;
use std::{
collections::{HashMap, HashSet},
fs,
io::Write,
process::Command,
rc::Rc,
sync::Arc,
};
use inkwell::{
use itertools::Itertools;
use parking_lot::{Mutex, RwLock};
use pyo3::{
create_exception, exceptions,
prelude::*,
types::{PyBytes, PyDict, PyNone, PySet},
};
use tempfile::{self, TempDir};
use nac3core::{
codegen::{
concrete_type::ConcreteTypeStore, gen_func_impl, irrt::load_irrt, CodeGenLLVMOptions,
CodeGenTargetMachineOptions, CodeGenTask, CodeGenerator, WithCall, WorkerRegistry,
},
inkwell::{
context::Context,
memory_buffer::MemoryBuffer,
module::{Linkage, Module},
module::{FlagBehavior, Linkage, Module},
passes::PassBuilderOptions,
support::is_multithreaded,
targets::*,
OptimizationLevel,
};
use itertools::Itertools;
use nac3core::codegen::{gen_func_impl, CodeGenLLVMOptions, CodeGenTargetMachineOptions};
use nac3core::toplevel::builtins::get_exn_constructor;
use nac3core::typecheck::typedef::{into_var_map, TypeEnum, Unifier, VarMap};
use nac3parser::{
ast::{ExprKind, Stmt, StmtKind, StrRef},
},
nac3parser::{
ast::{Constant, ExprKind, Located, Stmt, StmtKind, StrRef},
parser::parse_program,
};
use pyo3::create_exception;
use pyo3::prelude::*;
use pyo3::{exceptions, types::PyBytes, types::PyDict, types::PySet};
use parking_lot::{Mutex, RwLock};
use nac3core::{
codegen::irrt::load_irrt,
codegen::{concrete_type::ConcreteTypeStore, CodeGenTask, WithCall, WorkerRegistry},
},
symbol_resolver::SymbolResolver,
toplevel::{
builtins::get_exn_constructor,
composer::{BuiltinFuncCreator, BuiltinFuncSpec, ComposerConfig, TopLevelComposer},
DefinitionId, GenCall, TopLevelDef,
},
typecheck::typedef::{FunSignature, FuncArg},
typecheck::{type_inferencer::PrimitiveStore, typedef::Type},
typecheck::{
type_inferencer::PrimitiveStore,
typedef::{into_var_map, FunSignature, FuncArg, Type, TypeEnum, Unifier, VarMap},
},
};
use nac3ld::Linker;
use crate::{
codegen::{
use codegen::{
attributes_writeback, gen_core_log, gen_rtio_log, rpc_codegen_callback, ArtiqCodeGenerator,
},
symbol_resolver::{DeferredEvaluationStore, InnerResolver, PythonHelper, Resolver},
};
use tempfile::{self, TempDir};
use symbol_resolver::{DeferredEvaluationStore, InnerResolver, PythonHelper, Resolver};
use timeline::TimeFns;
mod codegen;
mod symbol_resolver;
mod timeline;
use timeline::TimeFns;
#[derive(PartialEq, Clone, Copy)]
enum Isa {
Host,
@ -147,14 +142,32 @@ impl Nac3 {
module: &PyObject,
registered_class_ids: &HashSet<u64>,
) -> PyResult<()> {
let (module_name, source_file) = Python::with_gil(|py| -> PyResult<(String, String)> {
let (module_name, source_file, source) =
Python::with_gil(|py| -> PyResult<(String, String, String)> {
let module: &PyAny = module.extract(py)?;
Ok((module.getattr("__name__")?.extract()?, module.getattr("__file__")?.extract()?))
let source_file = module.getattr("__file__");
let (source_file, source) = if let Ok(source_file) = source_file {
let source_file = source_file.extract()?;
(
source_file,
fs::read_to_string(source_file).map_err(|e| {
exceptions::PyIOError::new_err(format!(
"failed to read input file: {e}"
))
})?,
)
} else {
// kernels submitted by content have no file
// but still can provide source by StringLoader
let get_src_fn = module
.getattr("__loader__")?
.extract::<PyObject>()?
.getattr(py, "get_source")?;
("<expcontent>", get_src_fn.call1(py, (PyNone::get(py),))?.extract(py)?)
};
Ok((module.getattr("__name__")?.extract()?, source_file.to_string(), source))
})?;
let source = fs::read_to_string(&source_file).map_err(|e| {
exceptions::PyIOError::new_err(format!("failed to read input file: {e}"))
})?;
let parser_result = parse_program(&source, source_file.into())
.map_err(|e| exceptions::PySyntaxError::new_err(format!("parse error: {e}")))?;
@ -194,10 +207,8 @@ impl Nac3 {
body.retain(|stmt| {
if let StmtKind::FunctionDef { ref decorator_list, .. } = stmt.node {
decorator_list.iter().any(|decorator| {
if let ExprKind::Name { id, .. } = decorator.node {
id.to_string() == "kernel"
|| id.to_string() == "portable"
|| id.to_string() == "rpc"
if let Some(id) = decorator_id_string(decorator) {
id == "kernel" || id == "portable" || id == "rpc"
} else {
false
}
@ -210,9 +221,8 @@ impl Nac3 {
}
StmtKind::FunctionDef { ref decorator_list, .. } => {
decorator_list.iter().any(|decorator| {
if let ExprKind::Name { id, .. } = decorator.node {
let id = id.to_string();
id == "extern" || id == "portable" || id == "kernel" || id == "rpc"
if let Some(id) = decorator_id_string(decorator) {
id == "extern" || id == "kernel" || id == "portable" || id == "rpc"
} else {
false
}
@ -448,7 +458,6 @@ impl Nac3 {
pyid_to_type: pyid_to_type.clone(),
primitive_ids: self.primitive_ids.clone(),
global_value_ids: global_value_ids.clone(),
class_names: Mutex::default(),
name_to_pyid: name_to_pyid.clone(),
module: module.clone(),
id_to_pyval: RwLock::default(),
@ -479,9 +488,25 @@ impl Nac3 {
match &stmt.node {
StmtKind::FunctionDef { decorator_list, .. } => {
if decorator_list.iter().any(|decorator| matches!(decorator.node, ExprKind::Name { id, .. } if id == "rpc".into())) {
store_fun.call1(py, (def_id.0.into_py(py), module.getattr(py, name.to_string().as_str()).unwrap())).unwrap();
rpc_ids.push((None, def_id));
if decorator_list
.iter()
.any(|decorator| decorator_id_string(decorator) == Some("rpc".to_string()))
{
store_fun
.call1(
py,
(
def_id.0.into_py(py),
module.getattr(py, name.to_string().as_str()).unwrap(),
),
)
.unwrap();
let is_async = decorator_list.iter().any(|decorator| {
decorator_get_flags(decorator)
.iter()
.any(|constant| *constant == Constant::Str("async".into()))
});
rpc_ids.push((None, def_id, is_async));
}
}
StmtKind::ClassDef { name, body, .. } => {
@ -489,19 +514,26 @@ impl Nac3 {
let class_obj = module.getattr(py, class_name.as_str()).unwrap();
for stmt in body {
if let StmtKind::FunctionDef { name, decorator_list, .. } = &stmt.node {
if decorator_list.iter().any(|decorator| matches!(decorator.node, ExprKind::Name { id, .. } if id == "rpc".into())) {
if decorator_list.iter().any(|decorator| {
decorator_id_string(decorator) == Some("rpc".to_string())
}) {
let is_async = decorator_list.iter().any(|decorator| {
decorator_get_flags(decorator)
.iter()
.any(|constant| *constant == Constant::Str("async".into()))
});
if name == &"__init__".into() {
return Err(CompileError::new_err(format!(
"compilation failed\n----------\nThe constructor of class {} should not be decorated with rpc decorator (at {})",
class_name, stmt.location
)));
}
rpc_ids.push((Some((class_obj.clone(), *name)), def_id));
rpc_ids.push((Some((class_obj.clone(), *name)), def_id, is_async));
}
}
}
}
_ => ()
_ => (),
}
let id = *name_to_pyid.get(&name).unwrap();
@ -540,13 +572,12 @@ impl Nac3 {
pyid_to_type: pyid_to_type.clone(),
primitive_ids: self.primitive_ids.clone(),
global_value_ids: global_value_ids.clone(),
class_names: Mutex::default(),
id_to_pyval: RwLock::default(),
id_to_primitive: RwLock::default(),
field_to_val: RwLock::default(),
name_to_pyid,
module: module.to_object(py),
helper,
helper: helper.clone(),
string_store: self.string_store.clone(),
exception_ids: self.exception_ids.clone(),
deferred_eval_store: self.deferred_eval_store.clone(),
@ -558,7 +589,7 @@ impl Nac3 {
.unwrap();
// Process IRRT
let context = inkwell::context::Context::create();
let context = Context::create();
let irrt = load_irrt(&context, resolver.as_ref());
let fun_signature =
@ -598,13 +629,12 @@ impl Nac3 {
let top_level = Arc::new(composer.make_top_level_context());
{
let rpc_codegen = rpc_codegen_callback();
let defs = top_level.definitions.read();
for (class_data, id) in &rpc_ids {
for (class_data, id, is_async) in &rpc_ids {
let mut def = defs[id.0].write();
match &mut *def {
TopLevelDef::Function { codegen_callback, .. } => {
*codegen_callback = Some(rpc_codegen.clone());
*codegen_callback = Some(rpc_codegen_callback(*is_async));
}
TopLevelDef::Class { methods, .. } => {
let (class_def, method_name) = class_data.as_ref().unwrap();
@ -615,7 +645,7 @@ impl Nac3 {
if let TopLevelDef::Function { codegen_callback, .. } =
&mut *defs[id.0].write()
{
*codegen_callback = Some(rpc_codegen.clone());
*codegen_callback = Some(rpc_codegen_callback(*is_async));
store_fun
.call1(
py,
@ -630,6 +660,11 @@ impl Nac3 {
}
}
}
TopLevelDef::Variable { .. } => {
return Err(CompileError::new_err(String::from(
"Unsupported @rpc annotation on global variable",
)))
}
}
}
}
@ -650,33 +685,12 @@ impl Nac3 {
let task = CodeGenTask {
subst: Vec::default(),
symbol_name: "__modinit__".to_string(),
body: instance.body,
signature,
resolver: resolver.clone(),
store,
unifier_index: instance.unifier_id,
calls: instance.calls,
id: 0,
};
let mut store = ConcreteTypeStore::new();
let mut cache = HashMap::new();
let signature = store.from_signature(
&mut composer.unifier,
&self.primitive,
&fun_signature,
&mut cache,
);
let signature = store.add_cty(signature);
let attributes_writeback_task = CodeGenTask {
subst: Vec::default(),
symbol_name: "attributes_writeback".to_string(),
body: Arc::new(Vec::default()),
signature,
resolver,
store,
unifier_index: instance.unifier_id,
calls: Arc::new(HashMap::default()),
calls: instance.calls,
id: 0,
};
@ -689,7 +703,7 @@ impl Nac3 {
let buffer = buffer.as_slice().into();
membuffer.lock().push(buffer);
})));
let size_t = Context::create()
let size_t = context
.ptr_sized_int_type(&self.get_llvm_target_machine().get_target_data(), None)
.get_bit_width();
let num_threads = if is_multithreaded() { 4 } else { 1 };
@ -700,19 +714,27 @@ impl Nac3 {
.collect();
let membuffer = membuffers.clone();
let mut has_return = false;
py.allow_threads(|| {
let (registry, handles) =
WorkerRegistry::create_workers(threads, top_level.clone(), &self.llvm_options, &f);
registry.add_task(task);
registry.wait_tasks_complete(handles);
let mut generator =
ArtiqCodeGenerator::new("attributes_writeback".to_string(), size_t, self.time_fns);
let context = inkwell::context::Context::create();
let module = context.create_module("attributes_writeback");
let mut generator = ArtiqCodeGenerator::new("main".to_string(), size_t, self.time_fns);
let context = Context::create();
let module = context.create_module("main");
let target_machine = self.llvm_options.create_target_machine().unwrap();
module.set_data_layout(&target_machine.get_target_data().get_data_layout());
module.set_triple(&target_machine.get_triple());
module.add_basic_value_flag(
"Debug Info Version",
FlagBehavior::Warning,
context.i32_type().const_int(3, false),
);
module.add_basic_value_flag(
"Dwarf Version",
FlagBehavior::Warning,
context.i32_type().const_int(4, false),
);
let builder = context.create_builder();
let (_, module, _) = gen_func_impl(
&context,
@ -720,9 +742,27 @@ impl Nac3 {
&registry,
builder,
module,
attributes_writeback_task,
task,
|generator, ctx| {
attributes_writeback(ctx, generator, inner_resolver.as_ref(), &host_attributes)
assert_eq!(instance.body.len(), 1, "toplevel module should have 1 statement");
let StmtKind::Expr { value: ref expr, .. } = instance.body[0].node else {
unreachable!("toplevel statement must be an expression")
};
let ExprKind::Call { .. } = expr.node else {
unreachable!("toplevel expression must be a function call")
};
let return_obj =
generator.gen_expr(ctx, expr)?.map(|value| (expr.custom.unwrap(), value));
has_return = return_obj.is_some();
registry.wait_tasks_complete(handles);
attributes_writeback(
ctx,
generator,
inner_resolver.as_ref(),
&host_attributes,
return_obj,
)
},
)
.unwrap();
@ -731,35 +771,23 @@ impl Nac3 {
membuffer.lock().push(buffer);
});
embedding_map.setattr("expects_return", has_return).unwrap();
// Link all modules into `main`.
let buffers = membuffers.lock();
let main = context
.create_module_from_ir(MemoryBuffer::create_from_memory_range(&buffers[0], "main"))
.create_module_from_ir(MemoryBuffer::create_from_memory_range(
buffers.last().unwrap(),
"main",
))
.unwrap();
for buffer in buffers.iter().skip(1) {
for buffer in buffers.iter().rev().skip(1) {
let other = context
.create_module_from_ir(MemoryBuffer::create_from_memory_range(buffer, "main"))
.unwrap();
main.link_in_module(other).map_err(|err| CompileError::new_err(err.to_string()))?;
}
let builder = context.create_builder();
let modinit_return = main
.get_function("__modinit__")
.unwrap()
.get_last_basic_block()
.unwrap()
.get_terminator()
.unwrap();
builder.position_before(&modinit_return);
builder
.build_call(
main.get_function("attributes_writeback").unwrap(),
&[],
"attributes_writeback",
)
.unwrap();
main.link_in_module(irrt).map_err(|err| CompileError::new_err(err.to_string()))?;
let mut function_iter = main.get_first_function();
@ -794,6 +822,20 @@ impl Nac3 {
panic!("Failed to run optimization for module `main`: {}", err.to_string());
}
Python::with_gil(|py| {
let string_store = self.string_store.read();
let mut string_store_vec = string_store.iter().collect::<Vec<_>>();
string_store_vec.sort_by(|(_s1, key1), (_s2, key2)| key1.cmp(key2));
for (s, key) in string_store_vec {
let embed_key: i32 = helper.store_str.call1(py, (s,)).unwrap().extract(py).unwrap();
assert_eq!(
embed_key, *key,
"string {s} is out of sync between embedding map (key={embed_key}) and \
the internal string store (key={key})"
);
}
});
link_fn(&main)
}
@ -846,6 +888,41 @@ impl Nac3 {
}
}
/// Retrieves the Name.id from a decorator, supports decorators with arguments.
fn decorator_id_string(decorator: &Located<ExprKind>) -> Option<String> {
if let ExprKind::Name { id, .. } = decorator.node {
// Bare decorator
return Some(id.to_string());
} else if let ExprKind::Call { func, .. } = &decorator.node {
// Decorators that are calls (e.g. "@rpc()") have Call for the node,
// need to extract the id from within.
if let ExprKind::Name { id, .. } = func.node {
return Some(id.to_string());
}
}
None
}
/// Retrieves flags from a decorator, if any.
fn decorator_get_flags(decorator: &Located<ExprKind>) -> Vec<Constant> {
let mut flags = vec![];
if let ExprKind::Call { keywords, .. } = &decorator.node {
for keyword in keywords {
if keyword.node.arg != Some("flags".into()) {
continue;
}
if let ExprKind::Set { elts } = &keyword.node.value.node {
for elt in elts {
if let ExprKind::Constant { value, .. } = &elt.node {
flags.push(value.clone());
}
}
}
}
}
flags
}
fn link_with_lld(elf_filename: String, obj_filename: String) -> PyResult<()> {
let linker_args = vec![
"-shared".to_string(),
@ -1008,6 +1085,48 @@ impl Nac3 {
let working_directory = tempfile::Builder::new().prefix("nac3-").tempdir().unwrap();
fs::write(working_directory.path().join("kernel.ld"), include_bytes!("kernel.ld")).unwrap();
let mut string_store: HashMap<String, i32> = HashMap::default();
// Keep this list of exceptions in sync with `EXCEPTION_ID_LOOKUP` in `artiq::firmware::ksupport::eh_artiq`
// The exceptions declared here must be defined in `artiq.coredevice.exceptions`
// Verify synchronization by running the test cases in `artiq.test.coredevice.test_exceptions`
let runtime_exception_names = [
"RTIOUnderflow",
"RTIOOverflow",
"RTIODestinationUnreachable",
"DMAError",
"I2CError",
"CacheError",
"SPIError",
"SubkernelError",
"0:AssertionError",
"0:AttributeError",
"0:IndexError",
"0:IOError",
"0:KeyError",
"0:NotImplementedError",
"0:OverflowError",
"0:RuntimeError",
"0:TimeoutError",
"0:TypeError",
"0:ValueError",
"0:ZeroDivisionError",
"0:LinAlgError",
"UnwrapNoneError",
];
// Preallocate runtime exception names
for (i, name) in runtime_exception_names.iter().enumerate() {
let exn_name = if name.find(':').is_none() {
format!("0:artiq.coredevice.exceptions.{name}")
} else {
(*name).to_string()
};
let id = i32::try_from(i).unwrap();
string_store.insert(exn_name, id);
}
Ok(Nac3 {
isa,
time_fns,
@ -1017,7 +1136,7 @@ impl Nac3 {
top_levels: Vec::default(),
pyid_to_def: Arc::default(),
working_directory,
string_store: Arc::default(),
string_store: Arc::new(string_store.into()),
exception_ids: Arc::default(),
deferred_eval_store: DeferredEvaluationStore::new(),
llvm_options: CodeGenLLVMOptions {
@ -1027,7 +1146,12 @@ impl Nac3 {
})
}
fn analyze(&mut self, functions: &PySet, classes: &PySet) -> PyResult<()> {
fn analyze(
&mut self,
functions: &PySet,
classes: &PySet,
content_modules: &PySet,
) -> PyResult<()> {
let (modules, class_ids) =
Python::with_gil(|py| -> PyResult<(HashMap<u64, PyObject>, HashSet<u64>)> {
let mut modules: HashMap<u64, PyObject> = HashMap::new();
@ -1037,14 +1161,22 @@ impl Nac3 {
let getmodule_fn = PyModule::import(py, "inspect")?.getattr("getmodule")?;
for function in functions {
let module = getmodule_fn.call1((function,))?.extract()?;
let module: PyObject = getmodule_fn.call1((function,))?.extract()?;
if !module.is_none(py) {
modules.insert(id_fn.call1((&module,))?.extract()?, module);
}
}
for class in classes {
let module = getmodule_fn.call1((class,))?.extract()?;
let module: PyObject = getmodule_fn.call1((class,))?.extract()?;
if !module.is_none(py) {
modules.insert(id_fn.call1((&module,))?.extract()?, module);
}
class_ids.insert(id_fn.call1((class,))?.extract()?);
}
for module in content_modules {
let module: PyObject = module.extract()?;
modules.insert(id_fn.call1((&module,))?.extract()?, module);
}
Ok((modules, class_ids))
})?;

View File

@ -1,16 +1,32 @@
use crate::PrimitivePythonId;
use inkwell::{
use std::{
collections::{HashMap, HashSet},
sync::{
atomic::{AtomicBool, Ordering::Relaxed},
Arc,
},
};
use itertools::Itertools;
use parking_lot::RwLock;
use pyo3::{
types::{PyDict, PyTuple},
PyAny, PyErr, PyObject, PyResult, Python,
};
use super::PrimitivePythonId;
use nac3core::{
codegen::{
types::{ndarray::NDArrayType, ProxyType},
values::ndarray::make_contiguous_strides,
CodeGenContext, CodeGenerator,
},
inkwell::{
module::Linkage,
types::{BasicType, BasicTypeEnum},
values::BasicValueEnum,
AddressSpace,
};
use itertools::Itertools;
use nac3core::{
codegen::{
classes::{NDArrayType, ProxyType},
CodeGenContext, CodeGenerator,
},
nac3parser::ast::{self, StrRef},
symbol_resolver::{StaticValue, SymbolResolver, SymbolValue, ValueEnum},
toplevel::{
helper::PrimDef,
@ -22,19 +38,6 @@ use nac3core::{
typedef::{into_var_map, iter_type_vars, Type, TypeEnum, TypeVar, Unifier, VarMap},
},
};
use nac3parser::ast::{self, StrRef};
use parking_lot::{Mutex, RwLock};
use pyo3::{
types::{PyDict, PyTuple},
PyAny, PyObject, PyResult, Python,
};
use std::{
collections::{HashMap, HashSet},
sync::{
atomic::{AtomicBool, Ordering::Relaxed},
Arc,
},
};
pub enum PrimitiveValue {
I32(i32),
@ -79,7 +82,6 @@ pub struct InnerResolver {
pub id_to_primitive: RwLock<HashMap<u64, PrimitiveValue>>,
pub field_to_val: RwLock<HashMap<ResolverField, Option<PyFieldHandle>>>,
pub global_value_ids: Arc<RwLock<HashMap<u64, PyObject>>>,
pub class_names: Mutex<HashMap<StrRef, Type>>,
pub pyid_to_def: Arc<RwLock<HashMap<u64, DefinitionId>>>,
pub pyid_to_type: Arc<RwLock<HashMap<u64, Type>>>,
pub primitive_ids: PrimitivePythonId,
@ -1083,18 +1085,19 @@ impl InnerResolver {
} else {
unreachable!("must be ndarray")
};
let (ndarray_dtype, ndarray_ndims) =
unpack_ndarray_var_tys(&mut ctx.unifier, ndarray_ty);
let (ndarray_dtype, _) = unpack_ndarray_var_tys(&mut ctx.unifier, ndarray_ty);
let llvm_i8 = ctx.ctx.i8_type();
let llvm_pi8 = llvm_i8.ptr_type(AddressSpace::default());
let llvm_usize = generator.get_size_type(ctx.ctx);
let ndarray_dtype_llvm_ty = ctx.get_llvm_type(generator, ndarray_dtype);
let ndarray_llvm_ty = NDArrayType::new(generator, ctx.ctx, ndarray_dtype_llvm_ty);
let llvm_ndarray = NDArrayType::from_unifier_type(generator, ctx, ndarray_ty);
let dtype = llvm_ndarray.element_type();
{
if self.global_value_ids.read().contains_key(&id) {
let global = ctx.module.get_global(&id_str).unwrap_or_else(|| {
ctx.module.add_global(
ndarray_llvm_ty.as_underlying_type(),
llvm_ndarray.as_base_type().get_element_type().into_struct_type(),
Some(AddressSpace::default()),
&id_str,
)
@ -1104,40 +1107,41 @@ impl InnerResolver {
self.global_value_ids.write().insert(id, obj.into());
}
let TypeEnum::TLiteral { values, .. } = &*ctx.unifier.get_ty_immutable(ndarray_ndims)
else {
unreachable!("Expected Literal for ndarray_ndims")
};
let ndarray_ndims = if values.len() == 1 {
values[0].clone()
} else {
todo!("Unpacking literal of more than one element unimplemented")
};
let Ok(ndarray_ndims) = u64::try_from(ndarray_ndims) else {
unreachable!("Expected u64 value for ndarray_ndims")
};
let ndims = llvm_ndarray.ndims().unwrap();
// Obtain the shape of the ndarray
let shape_tuple: &PyTuple = obj.getattr("shape")?.downcast()?;
assert_eq!(shape_tuple.len(), ndarray_ndims as usize);
let shape_values: Result<Option<Vec<_>>, _> = shape_tuple
assert_eq!(shape_tuple.len(), ndims as usize);
// The Rust type inferencer cannot figure this out
let shape_values = shape_tuple
.iter()
.enumerate()
.map(|(i, elem)| {
self.get_obj_value(py, elem, ctx, generator, ctx.primitives.usize()).map_err(
|e| super::CompileError::new_err(format!("Error getting element {i}: {e}")),
)
let value = self
.get_obj_value(py, elem, ctx, generator, ctx.primitives.usize())
.map_err(|e| {
super::CompileError::new_err(format!("Error getting element {i}: {e}"))
})?
.unwrap();
let value = value.into_int_value();
Ok(value)
})
.collect();
let shape_values = shape_values?.unwrap();
let shape_values = llvm_usize.const_array(
&shape_values.into_iter().map(BasicValueEnum::into_int_value).collect_vec(),
);
.collect::<Result<Vec<_>, PyErr>>()?;
// Also use this opportunity to get the constant values of `shape_values` for calculating strides.
let shape_u64s = shape_values
.iter()
.map(|dim| {
assert!(dim.is_const());
dim.get_zero_extended_constant().unwrap()
})
.collect_vec();
let shape_values = llvm_usize.const_array(&shape_values);
// create a global for ndarray.shape and initialize it using the shape
let shape_global = ctx.module.add_global(
llvm_usize.array_type(ndarray_ndims as u32),
llvm_usize.array_type(ndims as u32),
Some(AddressSpace::default()),
&(id_str.clone() + ".shape"),
);
@ -1145,17 +1149,25 @@ impl InnerResolver {
// Obtain the (flattened) elements of the ndarray
let sz: usize = obj.getattr("size")?.extract()?;
let data: Result<Option<Vec<_>>, _> = (0..sz)
let data: Vec<_> = (0..sz)
.map(|i| {
obj.getattr("flat")?.get_item(i).and_then(|elem| {
self.get_obj_value(py, elem, ctx, generator, ndarray_dtype).map_err(|e| {
super::CompileError::new_err(format!("Error getting element {i}: {e}"))
let value = self
.get_obj_value(py, elem, ctx, generator, ndarray_dtype)
.map_err(|e| {
super::CompileError::new_err(format!(
"Error getting element {i}: {e}"
))
})?
.unwrap();
assert_eq!(value.get_type(), dtype);
Ok(value)
})
})
})
.collect();
let data = data?.unwrap().into_iter();
let data = match ndarray_dtype_llvm_ty {
.try_collect()?;
let data = data.into_iter();
let data = match dtype {
BasicTypeEnum::ArrayType(ty) => {
ty.const_array(&data.map(BasicValueEnum::into_array_value).collect_vec())
}
@ -1180,34 +1192,68 @@ impl InnerResolver {
};
// create a global for ndarray.data and initialize it using the elements
//
// NOTE: NDArray's `data` is `u8*`. Here, `data_global` is an array of `dtype`.
// We will have to cast it to an `u8*` later.
let data_global = ctx.module.add_global(
ndarray_dtype_llvm_ty.array_type(sz as u32),
dtype.array_type(sz as u32),
Some(AddressSpace::default()),
&(id_str.clone() + ".data"),
);
data_global.set_initializer(&data);
// Get the constant itemsize.
let itemsize = dtype.size_of().unwrap();
let itemsize = itemsize.get_zero_extended_constant().unwrap();
// Create the strides needed for ndarray.strides
let strides = make_contiguous_strides(itemsize, ndims, &shape_u64s);
let strides =
strides.into_iter().map(|stride| llvm_usize.const_int(stride, false)).collect_vec();
let strides = llvm_usize.const_array(&strides);
// create a global for ndarray.strides and initialize it
let strides_global = ctx.module.add_global(
llvm_i8.array_type(ndims as u32),
Some(AddressSpace::default()),
&format!("${id_str}.strides"),
);
strides_global.set_initializer(&strides);
// create a global for the ndarray object and initialize it
let value = ndarray_llvm_ty.as_underlying_type().const_named_struct(&[
llvm_usize.const_int(ndarray_ndims, false).into(),
shape_global
.as_pointer_value()
.const_cast(llvm_usize.ptr_type(AddressSpace::default()))
.into(),
data_global
.as_pointer_value()
.const_cast(ndarray_dtype_llvm_ty.ptr_type(AddressSpace::default()))
.into(),
// NOTE: data_global is an array of dtype, we want a `u8*`.
let ndarray_data = data_global.as_pointer_value();
let ndarray_data = ctx.builder.build_pointer_cast(ndarray_data, llvm_pi8, "").unwrap();
let ndarray_itemsize = llvm_usize.const_int(itemsize, false);
let ndarray_ndims = llvm_usize.const_int(ndims, false);
let ndarray_shape = shape_global.as_pointer_value();
let ndarray_strides = strides_global.as_pointer_value();
let ndarray = llvm_ndarray
.as_base_type()
.get_element_type()
.into_struct_type()
.const_named_struct(&[
ndarray_itemsize.into(),
ndarray_ndims.into(),
ndarray_shape.into(),
ndarray_strides.into(),
ndarray_data.into(),
]);
let ndarray = ctx.module.add_global(
ndarray_llvm_ty.as_underlying_type(),
let ndarray_global = ctx.module.add_global(
llvm_ndarray.as_base_type().get_element_type().into_struct_type(),
Some(AddressSpace::default()),
&id_str,
);
ndarray.set_initializer(&value);
ndarray_global.set_initializer(&ndarray);
Ok(Some(ndarray.as_pointer_value().into()))
Ok(Some(ndarray_global.as_pointer_value().into()))
} else if ty_id == self.primitive_ids.tuple {
let expected_ty_enum = ctx.unifier.get_ty_immutable(expected_ty);
let TypeEnum::TTuple { ty, is_vararg_ctx: false } = expected_ty_enum.as_ref() else {
@ -1468,6 +1514,7 @@ impl SymbolResolver for Resolver {
&self,
id: StrRef,
_: &mut CodeGenContext<'ctx, '_>,
_: &mut dyn CodeGenerator,
) -> Option<ValueEnum<'ctx>> {
let sym_value = {
let id_to_val = self.0.id_to_pyval.read();
@ -1529,10 +1576,7 @@ impl SymbolResolver for Resolver {
if let Some(id) = string_store.get(s) {
*id
} else {
let id = Python::with_gil(|py| -> PyResult<i32> {
self.0.helper.store_str.call1(py, (s,))?.extract(py)
})
.unwrap();
let id = i32::try_from(string_store.len()).unwrap();
string_store.insert(s.into(), id);
id
}

View File

@ -1,9 +1,12 @@
use inkwell::{
use itertools::Either;
use nac3core::{
codegen::CodeGenContext,
inkwell::{
values::{BasicValueEnum, CallSiteValue},
AddressSpace, AtomicOrdering,
},
};
use itertools::Either;
use nac3core::codegen::CodeGenContext;
/// Functions for manipulating the timeline.
pub trait TimeFns {
@ -31,7 +34,7 @@ impl TimeFns for NowPinningTimeFns64 {
.unwrap_or_else(|| ctx.module.add_global(i64_type, None, "now"));
let now_hiptr = ctx
.builder
.build_bitcast(now, i32_type.ptr_type(AddressSpace::default()), "now.hi.addr")
.build_bit_cast(now, i32_type.ptr_type(AddressSpace::default()), "now.hi.addr")
.map(BasicValueEnum::into_pointer_value)
.unwrap();
@ -80,7 +83,7 @@ impl TimeFns for NowPinningTimeFns64 {
.unwrap_or_else(|| ctx.module.add_global(i64_type, None, "now"));
let now_hiptr = ctx
.builder
.build_bitcast(now, i32_type.ptr_type(AddressSpace::default()), "now.hi.addr")
.build_bit_cast(now, i32_type.ptr_type(AddressSpace::default()), "now.hi.addr")
.map(BasicValueEnum::into_pointer_value)
.unwrap();
@ -109,7 +112,7 @@ impl TimeFns for NowPinningTimeFns64 {
.unwrap_or_else(|| ctx.module.add_global(i64_type, None, "now"));
let now_hiptr = ctx
.builder
.build_bitcast(now, i32_type.ptr_type(AddressSpace::default()), "now.hi.addr")
.build_bit_cast(now, i32_type.ptr_type(AddressSpace::default()), "now.hi.addr")
.map(BasicValueEnum::into_pointer_value)
.unwrap();
@ -207,7 +210,7 @@ impl TimeFns for NowPinningTimeFns {
.unwrap_or_else(|| ctx.module.add_global(i64_type, None, "now"));
let now_hiptr = ctx
.builder
.build_bitcast(now, i32_type.ptr_type(AddressSpace::default()), "now.hi.addr")
.build_bit_cast(now, i32_type.ptr_type(AddressSpace::default()), "now.hi.addr")
.map(BasicValueEnum::into_pointer_value)
.unwrap();
@ -258,7 +261,7 @@ impl TimeFns for NowPinningTimeFns {
let time_lo = ctx.builder.build_int_truncate(time, i32_type, "time.lo").unwrap();
let now_hiptr = ctx
.builder
.build_bitcast(now, i32_type.ptr_type(AddressSpace::default()), "now.hi.addr")
.build_bit_cast(now, i32_type.ptr_type(AddressSpace::default()), "now.hi.addr")
.map(BasicValueEnum::into_pointer_value)
.unwrap();

View File

@ -10,7 +10,6 @@ constant-optimization = ["fold"]
fold = []
[dependencies]
lazy_static = "1.5"
parking_lot = "0.12"
string-interner = "0.17"
fxhash = "0.2"

View File

@ -5,14 +5,12 @@ pub use crate::location::Location;
use fxhash::FxBuildHasher;
use parking_lot::{Mutex, MutexGuard};
use std::{cell::RefCell, collections::HashMap, fmt};
use std::{cell::RefCell, collections::HashMap, fmt, sync::LazyLock};
use string_interner::{symbol::SymbolU32, DefaultBackend, StringInterner};
pub type Interner = StringInterner<DefaultBackend, FxBuildHasher>;
lazy_static! {
static ref INTERNER: Mutex<Interner> =
Mutex::new(StringInterner::with_hasher(FxBuildHasher::default()));
}
static INTERNER: LazyLock<Mutex<Interner>> =
LazyLock::new(|| Mutex::new(StringInterner::with_hasher(FxBuildHasher::default())));
thread_local! {
static LOCAL_INTERNER: RefCell<HashMap<String, StrRef>> = RefCell::default();

View File

@ -1,10 +1,4 @@
#![deny(
future_incompatible,
let_underscore,
nonstandard_style,
rust_2024_compatibility,
clippy::all
)]
#![deny(future_incompatible, let_underscore, nonstandard_style, clippy::all)]
#![warn(clippy::pedantic)]
#![allow(
clippy::missing_errors_doc,
@ -14,9 +8,6 @@
clippy::wildcard_imports
)]
#[macro_use]
extern crate lazy_static;
mod ast_gen;
mod constant;
#[cfg(feature = "fold")]

View File

@ -5,22 +5,25 @@ authors = ["M-Labs"]
edition = "2021"
[features]
default = ["derive"]
derive = ["dep:nac3core_derive"]
no-escape-analysis = []
[dependencies]
itertools = "0.13"
crossbeam = "0.8"
indexmap = "2.2"
indexmap = "2.6"
parking_lot = "0.12"
rayon = "1.8"
rayon = "1.10"
nac3core_derive = { path = "nac3core_derive", optional = true }
nac3parser = { path = "../nac3parser" }
strum = "0.26"
strum_macros = "0.26"
[dependencies.inkwell]
version = "0.4"
version = "0.5"
default-features = false
features = ["llvm14-0", "target-x86", "target-arm", "target-riscv", "no-libffi-linking"]
features = ["llvm14-0-prefer-dynamic", "target-x86", "target-arm", "target-riscv", "no-libffi-linking"]
[dev-dependencies]
test-case = "1.2.0"

View File

@ -1,4 +1,3 @@
use regex::Regex;
use std::{
env,
fs::File,
@ -7,6 +6,8 @@ use std::{
process::{Command, Stdio},
};
use regex::Regex;
fn main() {
let out_dir = env::var("OUT_DIR").unwrap();
let out_dir = Path::new(&out_dir);
@ -55,9 +56,8 @@ fn main() {
let output = Command::new("clang-irrt")
.args(flags)
.output()
.map(|o| {
.inspect(|o| {
assert!(o.status.success(), "{}", std::str::from_utf8(&o.stderr).unwrap());
o
})
.unwrap();

View File

@ -1,6 +1,10 @@
#include <irrt/exception.hpp>
#include <irrt/int_types.hpp>
#include <irrt/list.hpp>
#include <irrt/math.hpp>
#include <irrt/ndarray.hpp>
#include <irrt/slice.hpp>
#include "irrt/exception.hpp"
#include "irrt/list.hpp"
#include "irrt/math.hpp"
#include "irrt/ndarray.hpp"
#include "irrt/range.hpp"
#include "irrt/slice.hpp"
#include "irrt/ndarray/basic.hpp"
#include "irrt/ndarray/def.hpp"
#include "irrt/ndarray/iter.hpp"
#include "irrt/ndarray/indexing.hpp"

View File

@ -1,9 +1,9 @@
#pragma once
#include <irrt/int_types.hpp>
#include "irrt/int_types.hpp"
template<typename SizeT>
struct CSlice {
uint8_t* base;
void* base;
SizeT len;
};

View File

@ -1,12 +1,12 @@
#pragma once
#include <irrt/cslice.hpp>
#include <irrt/int_types.hpp>
#include "irrt/cslice.hpp"
#include "irrt/int_types.hpp"
/**
* @brief The int type of ARTIQ exception IDs.
*/
typedef int32_t ExceptionId;
using ExceptionId = int32_t;
/*
* Set of exceptions C++ IRRT can use.
@ -55,11 +55,14 @@ void _raise_exception_helper(ExceptionId id,
int64_t param2) {
Exception<SizeT> e = {
.id = id,
.filename = {.base = reinterpret_cast<const uint8_t*>(filename), .len = __builtin_strlen(filename)},
.filename = {.base = reinterpret_cast<void*>(const_cast<char*>(filename)),
.len = static_cast<SizeT>(__builtin_strlen(filename))},
.line = line,
.column = 0,
.function = {.base = reinterpret_cast<const uint8_t*>(function), .len = __builtin_strlen(function)},
.msg = {.base = reinterpret_cast<const uint8_t*>(msg), .len = __builtin_strlen(msg)},
.function = {.base = reinterpret_cast<void*>(const_cast<char*>(function)),
.len = static_cast<SizeT>(__builtin_strlen(function))},
.msg = {.base = reinterpret_cast<void*>(const_cast<char*>(msg)),
.len = static_cast<SizeT>(__builtin_strlen(msg))},
};
e.params[0] = param0;
e.params[1] = param1;
@ -67,6 +70,7 @@ void _raise_exception_helper(ExceptionId id,
__nac3_raise(reinterpret_cast<void*>(&e));
__builtin_unreachable();
}
} // namespace
/**
* @brief Raise an exception with location details (location in the IRRT source files).
@ -79,4 +83,3 @@ void _raise_exception_helper(ExceptionId id,
*/
#define raise_exception(SizeT, id, msg, param0, param1, param2) \
_raise_exception_helper<SizeT>(id, __FILE__, __LINE__, __FUNCTION__, msg, param0, param1, param2)
} // namespace

View File

@ -1,13 +1,27 @@
#pragma once
#if __STDC_VERSION__ >= 202000
using int8_t = _BitInt(8);
using uint8_t = unsigned _BitInt(8);
using int32_t = _BitInt(32);
using uint32_t = unsigned _BitInt(32);
using int64_t = _BitInt(64);
using uint64_t = unsigned _BitInt(64);
#else
#pragma clang diagnostic push
#pragma clang diagnostic ignored "-Wdeprecated-type"
using int8_t = _ExtInt(8);
using uint8_t = unsigned _ExtInt(8);
using int32_t = _ExtInt(32);
using uint32_t = unsigned _ExtInt(32);
using int64_t = _ExtInt(64);
using uint64_t = unsigned _ExtInt(64);
#pragma clang diagnostic pop
#endif
// NDArray indices are always `uint32_t`.
using NDIndex = uint32_t;
using NDIndexInt = uint32_t;
// The type of an index or a value describing the length of a range/slice is always `int32_t`.
using SliceIndex = int32_t;

View File

@ -1,7 +1,7 @@
#pragma once
#include <irrt/int_types.hpp>
#include <irrt/math_util.hpp>
#include "irrt/int_types.hpp"
#include "irrt/math_util.hpp"
extern "C" {
// Handle list assignment and dropping part of the list when
@ -13,12 +13,12 @@ extern "C" {
SliceIndex __nac3_list_slice_assign_var_size(SliceIndex dest_start,
SliceIndex dest_end,
SliceIndex dest_step,
uint8_t* dest_arr,
void* dest_arr,
SliceIndex dest_arr_len,
SliceIndex src_start,
SliceIndex src_end,
SliceIndex src_step,
uint8_t* src_arr,
void* src_arr,
SliceIndex src_arr_len,
const SliceIndex size) {
/* if dest_arr_len == 0, do nothing since we do not support extending list */
@ -29,11 +29,13 @@ SliceIndex __nac3_list_slice_assign_var_size(SliceIndex dest_start,
const SliceIndex src_len = (src_end >= src_start) ? (src_end - src_start + 1) : 0;
const SliceIndex dest_len = (dest_end >= dest_start) ? (dest_end - dest_start + 1) : 0;
if (src_len > 0) {
__builtin_memmove(dest_arr + dest_start * size, src_arr + src_start * size, src_len * size);
__builtin_memmove(static_cast<uint8_t*>(dest_arr) + dest_start * size,
static_cast<uint8_t*>(src_arr) + src_start * size, src_len * size);
}
if (dest_len > 0) {
/* dropping */
__builtin_memmove(dest_arr + (dest_start + src_len) * size, dest_arr + (dest_end + 1) * size,
__builtin_memmove(static_cast<uint8_t*>(dest_arr) + (dest_start + src_len) * size,
static_cast<uint8_t*>(dest_arr) + (dest_end + 1) * size,
(dest_arr_len - dest_end - 1) * size);
}
/* shrink size */
@ -44,7 +46,7 @@ SliceIndex __nac3_list_slice_assign_var_size(SliceIndex dest_start,
&& !(max(dest_start, dest_end) < min(src_start, src_end)
|| max(src_start, src_end) < min(dest_start, dest_end));
if (need_alloca) {
uint8_t* tmp = reinterpret_cast<uint8_t*>(__builtin_alloca(src_arr_len * size));
void* tmp = __builtin_alloca(src_arr_len * size);
__builtin_memcpy(tmp, src_arr, src_arr_len * size);
src_arr = tmp;
}
@ -53,20 +55,24 @@ SliceIndex __nac3_list_slice_assign_var_size(SliceIndex dest_start,
for (; (src_step > 0) ? (src_ind <= src_end) : (src_ind >= src_end); src_ind += src_step, dest_ind += dest_step) {
/* for constant optimization */
if (size == 1) {
__builtin_memcpy(dest_arr + dest_ind, src_arr + src_ind, 1);
__builtin_memcpy(static_cast<uint8_t*>(dest_arr) + dest_ind, static_cast<uint8_t*>(src_arr) + src_ind, 1);
} else if (size == 4) {
__builtin_memcpy(dest_arr + dest_ind * 4, src_arr + src_ind * 4, 4);
__builtin_memcpy(static_cast<uint8_t*>(dest_arr) + dest_ind * 4,
static_cast<uint8_t*>(src_arr) + src_ind * 4, 4);
} else if (size == 8) {
__builtin_memcpy(dest_arr + dest_ind * 8, src_arr + src_ind * 8, 8);
__builtin_memcpy(static_cast<uint8_t*>(dest_arr) + dest_ind * 8,
static_cast<uint8_t*>(src_arr) + src_ind * 8, 8);
} else {
/* memcpy for var size, cannot overlap after previous alloca */
__builtin_memcpy(dest_arr + dest_ind * size, src_arr + src_ind * size, size);
__builtin_memcpy(static_cast<uint8_t*>(dest_arr) + dest_ind * size,
static_cast<uint8_t*>(src_arr) + src_ind * size, size);
}
}
/* only dest_step == 1 can we shrink the dest list. */
/* size should be ensured prior to calling this function */
if (dest_step == 1 && dest_end >= dest_start) {
__builtin_memmove(dest_arr + dest_ind * size, dest_arr + (dest_end + 1) * size,
__builtin_memmove(static_cast<uint8_t*>(dest_arr) + dest_ind * size,
static_cast<uint8_t*>(dest_arr) + (dest_end + 1) * size,
(dest_arr_len - dest_end - 1) * size);
return dest_arr_len - (dest_end - dest_ind) - 1;
}

View File

@ -90,4 +90,4 @@ double __nac3_j0(double x) {
return j0(x);
}
}
} // namespace

View File

@ -1,6 +1,8 @@
#pragma once
#include <irrt/int_types.hpp>
#include "irrt/int_types.hpp"
// TODO: To be deleted since NDArray with strides is done.
namespace {
template<typename SizeT>
@ -17,7 +19,7 @@ SizeT __nac3_ndarray_calc_size_impl(const SizeT* list_data, SizeT list_len, Size
}
template<typename SizeT>
void __nac3_ndarray_calc_nd_indices_impl(SizeT index, const SizeT* dims, SizeT num_dims, NDIndex* idxs) {
void __nac3_ndarray_calc_nd_indices_impl(SizeT index, const SizeT* dims, SizeT num_dims, NDIndexInt* idxs) {
SizeT stride = 1;
for (SizeT dim = 0; dim < num_dims; dim++) {
SizeT i = num_dims - dim - 1;
@ -28,7 +30,10 @@ void __nac3_ndarray_calc_nd_indices_impl(SizeT index, const SizeT* dims, SizeT n
}
template<typename SizeT>
SizeT __nac3_ndarray_flatten_index_impl(const SizeT* dims, SizeT num_dims, const NDIndex* indices, SizeT num_indices) {
SizeT __nac3_ndarray_flatten_index_impl(const SizeT* dims,
SizeT num_dims,
const NDIndexInt* indices,
SizeT num_indices) {
SizeT idx = 0;
SizeT stride = 1;
for (SizeT i = 0; i < num_dims; ++i) {
@ -75,8 +80,8 @@ void __nac3_ndarray_calc_broadcast_impl(const SizeT* lhs_dims,
template<typename SizeT>
void __nac3_ndarray_calc_broadcast_idx_impl(const SizeT* src_dims,
SizeT src_ndims,
const NDIndex* in_idx,
NDIndex* out_idx) {
const NDIndexInt* in_idx,
NDIndexInt* out_idx) {
for (SizeT i = 0; i < src_ndims; ++i) {
SizeT src_i = src_ndims - i - 1;
out_idx[src_i] = src_dims[src_i] == 1 ? 0 : in_idx[src_i];
@ -94,21 +99,23 @@ __nac3_ndarray_calc_size64(const uint64_t* list_data, uint64_t list_len, uint64_
return __nac3_ndarray_calc_size_impl(list_data, list_len, begin_idx, end_idx);
}
void __nac3_ndarray_calc_nd_indices(uint32_t index, const uint32_t* dims, uint32_t num_dims, NDIndex* idxs) {
void __nac3_ndarray_calc_nd_indices(uint32_t index, const uint32_t* dims, uint32_t num_dims, NDIndexInt* idxs) {
__nac3_ndarray_calc_nd_indices_impl(index, dims, num_dims, idxs);
}
void __nac3_ndarray_calc_nd_indices64(uint64_t index, const uint64_t* dims, uint64_t num_dims, NDIndex* idxs) {
void __nac3_ndarray_calc_nd_indices64(uint64_t index, const uint64_t* dims, uint64_t num_dims, NDIndexInt* idxs) {
__nac3_ndarray_calc_nd_indices_impl(index, dims, num_dims, idxs);
}
uint32_t
__nac3_ndarray_flatten_index(const uint32_t* dims, uint32_t num_dims, const NDIndex* indices, uint32_t num_indices) {
__nac3_ndarray_flatten_index(const uint32_t* dims, uint32_t num_dims, const NDIndexInt* indices, uint32_t num_indices) {
return __nac3_ndarray_flatten_index_impl(dims, num_dims, indices, num_indices);
}
uint64_t
__nac3_ndarray_flatten_index64(const uint64_t* dims, uint64_t num_dims, const NDIndex* indices, uint64_t num_indices) {
uint64_t __nac3_ndarray_flatten_index64(const uint64_t* dims,
uint64_t num_dims,
const NDIndexInt* indices,
uint64_t num_indices) {
return __nac3_ndarray_flatten_index_impl(dims, num_dims, indices, num_indices);
}
@ -130,15 +137,15 @@ void __nac3_ndarray_calc_broadcast64(const uint64_t* lhs_dims,
void __nac3_ndarray_calc_broadcast_idx(const uint32_t* src_dims,
uint32_t src_ndims,
const NDIndex* in_idx,
NDIndex* out_idx) {
const NDIndexInt* in_idx,
NDIndexInt* out_idx) {
__nac3_ndarray_calc_broadcast_idx_impl(src_dims, src_ndims, in_idx, out_idx);
}
void __nac3_ndarray_calc_broadcast_idx64(const uint64_t* src_dims,
uint64_t src_ndims,
const NDIndex* in_idx,
NDIndex* out_idx) {
const NDIndexInt* in_idx,
NDIndexInt* out_idx) {
__nac3_ndarray_calc_broadcast_idx_impl(src_dims, src_ndims, in_idx, out_idx);
}
}
} // namespace

View File

@ -0,0 +1,342 @@
#pragma once
#include "irrt/debug.hpp"
#include "irrt/exception.hpp"
#include "irrt/int_types.hpp"
#include "irrt/ndarray/def.hpp"
namespace {
namespace ndarray {
namespace basic {
/**
* @brief Assert that `shape` does not contain negative dimensions.
*
* @param ndims Number of dimensions in `shape`
* @param shape The shape to check on
*/
template<typename SizeT>
void assert_shape_no_negative(SizeT ndims, const SizeT* shape) {
for (SizeT axis = 0; axis < ndims; axis++) {
if (shape[axis] < 0) {
raise_exception(SizeT, EXN_VALUE_ERROR,
"negative dimensions are not allowed; axis {0} "
"has dimension {1}",
axis, shape[axis], NO_PARAM);
}
}
}
/**
* @brief Assert that two shapes are the same in the context of writing output to an ndarray.
*/
template<typename SizeT>
void assert_output_shape_same(SizeT ndarray_ndims,
const SizeT* ndarray_shape,
SizeT output_ndims,
const SizeT* output_shape) {
if (ndarray_ndims != output_ndims) {
// There is no corresponding NumPy error message like this.
raise_exception(SizeT, EXN_VALUE_ERROR, "Cannot write output of ndims {0} to an ndarray with ndims {1}",
output_ndims, ndarray_ndims, NO_PARAM);
}
for (SizeT axis = 0; axis < ndarray_ndims; axis++) {
if (ndarray_shape[axis] != output_shape[axis]) {
// There is no corresponding NumPy error message like this.
raise_exception(SizeT, EXN_VALUE_ERROR,
"Mismatched dimensions on axis {0}, output has "
"dimension {1}, but destination ndarray has dimension {2}.",
axis, output_shape[axis], ndarray_shape[axis]);
}
}
}
/**
* @brief Return the number of elements of an ndarray given its shape.
*
* @param ndims Number of dimensions in `shape`
* @param shape The shape of the ndarray
*/
template<typename SizeT>
SizeT calc_size_from_shape(SizeT ndims, const SizeT* shape) {
SizeT size = 1;
for (SizeT axis = 0; axis < ndims; axis++)
size *= shape[axis];
return size;
}
/**
* @brief Compute the array indices of the `nth` (0-based) element of an ndarray given only its shape.
*
* @param ndims Number of elements in `shape` and `indices`
* @param shape The shape of the ndarray
* @param indices The returned indices indexing the ndarray with shape `shape`.
* @param nth The index of the element of interest.
*/
template<typename SizeT>
void set_indices_by_nth(SizeT ndims, const SizeT* shape, SizeT* indices, SizeT nth) {
for (SizeT i = 0; i < ndims; i++) {
SizeT axis = ndims - i - 1;
SizeT dim = shape[axis];
indices[axis] = nth % dim;
nth /= dim;
}
}
/**
* @brief Return the number of elements of an `ndarray`
*
* This function corresponds to `<an_ndarray>.size`
*/
template<typename SizeT>
SizeT size(const NDArray<SizeT>* ndarray) {
return calc_size_from_shape(ndarray->ndims, ndarray->shape);
}
/**
* @brief Return of the number of its content of an `ndarray`.
*
* This function corresponds to `<an_ndarray>.nbytes`.
*/
template<typename SizeT>
SizeT nbytes(const NDArray<SizeT>* ndarray) {
return size(ndarray) * ndarray->itemsize;
}
/**
* @brief Get the `len()` of an ndarray, and asserts that `ndarray` is a sized object.
*
* This function corresponds to `<an_ndarray>.__len__`.
*
* @param dst_length The length.
*/
template<typename SizeT>
SizeT len(const NDArray<SizeT>* ndarray) {
if (ndarray->ndims != 0) {
return ndarray->shape[0];
}
// numpy prohibits `__len__` on unsized objects
raise_exception(SizeT, EXN_TYPE_ERROR, "len() of unsized object", NO_PARAM, NO_PARAM, NO_PARAM);
__builtin_unreachable();
}
/**
* @brief Return a boolean indicating if `ndarray` is (C-)contiguous.
*
* You may want to see ndarray's rules for C-contiguity:
* https://github.com/numpy/numpy/blob/df256d0d2f3bc6833699529824781c58f9c6e697/numpy/core/src/multiarray/flagsobject.c#L95C1-L99C45
*/
template<typename SizeT>
bool is_c_contiguous(const NDArray<SizeT>* ndarray) {
// References:
// - tinynumpy's implementation:
// https://github.com/wadetb/tinynumpy/blob/0d23d22e07062ffab2afa287374c7b366eebdda1/tinynumpy/tinynumpy.py#L102
// - ndarray's flags["C_CONTIGUOUS"]:
// https://numpy.org/doc/stable/reference/generated/numpy.ndarray.flags.html#numpy.ndarray.flags
// - ndarray's rules for C-contiguity:
// https://github.com/numpy/numpy/blob/df256d0d2f3bc6833699529824781c58f9c6e697/numpy/core/src/multiarray/flagsobject.c#L95C1-L99C45
// From
// https://github.com/numpy/numpy/blob/df256d0d2f3bc6833699529824781c58f9c6e697/numpy/core/src/multiarray/flagsobject.c#L95C1-L99C45:
//
// The traditional rule is that for an array to be flagged as C contiguous,
// the following must hold:
//
// strides[-1] == itemsize
// strides[i] == shape[i+1] * strides[i + 1]
// [...]
// According to these rules, a 0- or 1-dimensional array is either both
// C- and F-contiguous, or neither; and an array with 2+ dimensions
// can be C- or F- contiguous, or neither, but not both. Though there
// there are exceptions for arrays with zero or one item, in the first
// case the check is relaxed up to and including the first dimension
// with shape[i] == 0. In the second case `strides == itemsize` will
// can be true for all dimensions and both flags are set.
if (ndarray->ndims == 0) {
return true;
}
if (ndarray->strides[ndarray->ndims - 1] != ndarray->itemsize) {
return false;
}
for (SizeT i = 1; i < ndarray->ndims; i++) {
SizeT axis_i = ndarray->ndims - i - 1;
if (ndarray->strides[axis_i] != ndarray->shape[axis_i + 1] * ndarray->strides[axis_i + 1]) {
return false;
}
}
return true;
}
/**
* @brief Return the pointer to the element indexed by `indices` along the ndarray's axes.
*
* This function does no bound check.
*/
template<typename SizeT>
void* get_pelement_by_indices(const NDArray<SizeT>* ndarray, const SizeT* indices) {
void* element = ndarray->data;
for (SizeT dim_i = 0; dim_i < ndarray->ndims; dim_i++)
element = static_cast<uint8_t*>(element) + indices[dim_i] * ndarray->strides[dim_i];
return element;
}
/**
* @brief Return the pointer to the nth (0-based) element of `ndarray` in flattened view.
*
* This function does no bound check.
*/
template<typename SizeT>
void* get_nth_pelement(const NDArray<SizeT>* ndarray, SizeT nth) {
void* element = ndarray->data;
for (SizeT i = 0; i < ndarray->ndims; i++) {
SizeT axis = ndarray->ndims - i - 1;
SizeT dim = ndarray->shape[axis];
element = static_cast<uint8_t*>(element) + ndarray->strides[axis] * (nth % dim);
nth /= dim;
}
return element;
}
/**
* @brief Update the strides of an ndarray given an ndarray `shape` to be contiguous.
*
* You might want to read https://ajcr.net/stride-guide-part-1/.
*/
template<typename SizeT>
void set_strides_by_shape(NDArray<SizeT>* ndarray) {
SizeT stride_product = 1;
for (SizeT i = 0; i < ndarray->ndims; i++) {
SizeT axis = ndarray->ndims - i - 1;
ndarray->strides[axis] = stride_product * ndarray->itemsize;
stride_product *= ndarray->shape[axis];
}
}
/**
* @brief Set an element in `ndarray`.
*
* @param pelement Pointer to the element in `ndarray` to be set.
* @param pvalue Pointer to the value `pelement` will be set to.
*/
template<typename SizeT>
void set_pelement_value(NDArray<SizeT>* ndarray, void* pelement, const void* pvalue) {
__builtin_memcpy(pelement, pvalue, ndarray->itemsize);
}
/**
* @brief Copy data from one ndarray to another of the exact same size and itemsize.
*
* Both ndarrays will be viewed in their flatten views when copying the elements.
*/
template<typename SizeT>
void copy_data(const NDArray<SizeT>* src_ndarray, NDArray<SizeT>* dst_ndarray) {
// TODO: Make this faster with memcpy when we see a contiguous segment.
// TODO: Handle overlapping.
debug_assert_eq(SizeT, src_ndarray->itemsize, dst_ndarray->itemsize);
for (SizeT i = 0; i < size(src_ndarray); i++) {
auto src_element = ndarray::basic::get_nth_pelement(src_ndarray, i);
auto dst_element = ndarray::basic::get_nth_pelement(dst_ndarray, i);
ndarray::basic::set_pelement_value(dst_ndarray, dst_element, src_element);
}
}
} // namespace basic
} // namespace ndarray
} // namespace
extern "C" {
using namespace ndarray::basic;
void __nac3_ndarray_util_assert_shape_no_negative(int32_t ndims, int32_t* shape) {
assert_shape_no_negative(ndims, shape);
}
void __nac3_ndarray_util_assert_shape_no_negative64(int64_t ndims, int64_t* shape) {
assert_shape_no_negative(ndims, shape);
}
void __nac3_ndarray_util_assert_output_shape_same(int32_t ndarray_ndims,
const int32_t* ndarray_shape,
int32_t output_ndims,
const int32_t* output_shape) {
assert_output_shape_same(ndarray_ndims, ndarray_shape, output_ndims, output_shape);
}
void __nac3_ndarray_util_assert_output_shape_same64(int64_t ndarray_ndims,
const int64_t* ndarray_shape,
int64_t output_ndims,
const int64_t* output_shape) {
assert_output_shape_same(ndarray_ndims, ndarray_shape, output_ndims, output_shape);
}
uint32_t __nac3_ndarray_size(NDArray<int32_t>* ndarray) {
return size(ndarray);
}
uint64_t __nac3_ndarray_size64(NDArray<int64_t>* ndarray) {
return size(ndarray);
}
uint32_t __nac3_ndarray_nbytes(NDArray<int32_t>* ndarray) {
return nbytes(ndarray);
}
uint64_t __nac3_ndarray_nbytes64(NDArray<int64_t>* ndarray) {
return nbytes(ndarray);
}
int32_t __nac3_ndarray_len(NDArray<int32_t>* ndarray) {
return len(ndarray);
}
int64_t __nac3_ndarray_len64(NDArray<int64_t>* ndarray) {
return len(ndarray);
}
bool __nac3_ndarray_is_c_contiguous(NDArray<int32_t>* ndarray) {
return is_c_contiguous(ndarray);
}
bool __nac3_ndarray_is_c_contiguous64(NDArray<int64_t>* ndarray) {
return is_c_contiguous(ndarray);
}
void* __nac3_ndarray_get_nth_pelement(const NDArray<int32_t>* ndarray, int32_t nth) {
return get_nth_pelement(ndarray, nth);
}
void* __nac3_ndarray_get_nth_pelement64(const NDArray<int64_t>* ndarray, int64_t nth) {
return get_nth_pelement(ndarray, nth);
}
void* __nac3_ndarray_get_pelement_by_indices(const NDArray<int32_t>* ndarray, int32_t* indices) {
return get_pelement_by_indices(ndarray, indices);
}
void* __nac3_ndarray_get_pelement_by_indices64(const NDArray<int64_t>* ndarray, int64_t* indices) {
return get_pelement_by_indices(ndarray, indices);
}
void __nac3_ndarray_set_strides_by_shape(NDArray<int32_t>* ndarray) {
set_strides_by_shape(ndarray);
}
void __nac3_ndarray_set_strides_by_shape64(NDArray<int64_t>* ndarray) {
set_strides_by_shape(ndarray);
}
void __nac3_ndarray_copy_data(NDArray<int32_t>* src_ndarray, NDArray<int32_t>* dst_ndarray) {
copy_data(src_ndarray, dst_ndarray);
}
void __nac3_ndarray_copy_data64(NDArray<int64_t>* src_ndarray, NDArray<int64_t>* dst_ndarray) {
copy_data(src_ndarray, dst_ndarray);
}
}

View File

@ -0,0 +1,51 @@
#pragma once
#include "irrt/int_types.hpp"
namespace {
/**
* @brief The NDArray object
*
* Official numpy implementation:
* https://github.com/numpy/numpy/blob/735a477f0bc2b5b84d0e72d92f224bde78d4e069/doc/source/reference/c-api/types-and-structures.rst#pyarrayinterface
*
* Note that this implementation is based on `PyArrayInterface` rather of `PyArrayObject`. The
* difference between `PyArrayInterface` and `PyArrayObject` (relevant to our implementation) is
* that `PyArrayInterface` *has* `itemsize` and uses `void*` for its `data`, whereas `PyArrayObject`
* does not require `itemsize` (probably using `strides[-1]` instead) and uses `char*` for its
* `data`. There are also minor differences in the struct layout.
*/
template<typename SizeT>
struct NDArray {
/**
* @brief The number of bytes of a single element in `data`.
*/
SizeT itemsize;
/**
* @brief The number of dimensions of this shape.
*/
SizeT ndims;
/**
* @brief The NDArray shape, with length equal to `ndims`.
*
* Note that it may contain 0.
*/
SizeT* shape;
/**
* @brief Array strides, with length equal to `ndims`
*
* The stride values are in units of bytes, not number of elements.
*
* Note that `strides` can have negative values or contain 0.
*/
SizeT* strides;
/**
* @brief The underlying data this `ndarray` is pointing to.
*/
void* data;
};
} // namespace

View File

@ -0,0 +1,220 @@
#pragma once
#include "irrt/exception.hpp"
#include "irrt/int_types.hpp"
#include "irrt/ndarray/basic.hpp"
#include "irrt/ndarray/def.hpp"
#include "irrt/range.hpp"
#include "irrt/slice.hpp"
namespace {
typedef uint8_t NDIndexType;
/**
* @brief A single element index
*
* `data` points to a `int32_t`.
*/
const NDIndexType ND_INDEX_TYPE_SINGLE_ELEMENT = 0;
/**
* @brief A slice index
*
* `data` points to a `Slice<int32_t>`.
*/
const NDIndexType ND_INDEX_TYPE_SLICE = 1;
/**
* @brief `np.newaxis` / `None`
*
* `data` is unused.
*/
const NDIndexType ND_INDEX_TYPE_NEWAXIS = 2;
/**
* @brief `Ellipsis` / `...`
*
* `data` is unused.
*/
const NDIndexType ND_INDEX_TYPE_ELLIPSIS = 3;
/**
* @brief An index used in ndarray indexing
*
* That is:
* ```
* my_ndarray[::-1, 3, ..., np.newaxis]
* ^^^^ ^ ^^^ ^^^^^^^^^^ each of these is represented by an NDIndex.
* ```
*/
struct NDIndex {
/**
* @brief Enum tag to specify the type of index.
*
* Please see the comment of each enum constant.
*/
NDIndexType type;
/**
* @brief The accompanying data associated with `type`.
*
* Please see the comment of each enum constant.
*/
uint8_t* data;
};
} // namespace
namespace {
namespace ndarray {
namespace indexing {
/**
* @brief Perform ndarray "basic indexing" (https://numpy.org/doc/stable/user/basics.indexing.html#basic-indexing)
*
* This function is very similar to performing `dst_ndarray = src_ndarray[indices]` in Python.
*
* This function also does proper assertions on `indices` to check for out of bounds access and more.
*
* # Notes on `dst_ndarray`
* The caller is responsible for allocating space for the resulting ndarray.
* Here is what this function expects from `dst_ndarray` when called:
* - `dst_ndarray->data` does not have to be initialized.
* - `dst_ndarray->itemsize` does not have to be initialized.
* - `dst_ndarray->ndims` must be initialized, and it must be equal to the expected `ndims` of the `dst_ndarray` after
* indexing `src_ndarray` with `indices`.
* - `dst_ndarray->shape` must be allocated, through it can contain uninitialized values.
* - `dst_ndarray->strides` must be allocated, through it can contain uninitialized values.
* When this function call ends:
* - `dst_ndarray->data` is set to `src_ndarray->data`.
* - `dst_ndarray->itemsize` is set to `src_ndarray->itemsize`.
* - `dst_ndarray->ndims` is unchanged.
* - `dst_ndarray->shape` is updated according to how `src_ndarray` is indexed.
* - `dst_ndarray->strides` is updated accordingly by how ndarray indexing works.
*
* @param indices indices to index `src_ndarray`, ordered in the same way you would write them in Python.
* @param src_ndarray The NDArray to be indexed.
* @param dst_ndarray The resulting NDArray after indexing. Further details in the comments above,
*/
template<typename SizeT>
void index(SizeT num_indices, const NDIndex* indices, const NDArray<SizeT>* src_ndarray, NDArray<SizeT>* dst_ndarray) {
// Validate `indices`.
// Expected value of `dst_ndarray->ndims`.
SizeT expected_dst_ndims = src_ndarray->ndims;
// To check for "too many indices for array: array is ?-dimensional, but ? were indexed"
SizeT num_indexed = 0;
// There may be ellipsis `...` in `indices`. There can only be 0 or 1 ellipsis.
SizeT num_ellipsis = 0;
for (SizeT i = 0; i < num_indices; i++) {
if (indices[i].type == ND_INDEX_TYPE_SINGLE_ELEMENT) {
expected_dst_ndims--;
num_indexed++;
} else if (indices[i].type == ND_INDEX_TYPE_SLICE) {
num_indexed++;
} else if (indices[i].type == ND_INDEX_TYPE_NEWAXIS) {
expected_dst_ndims++;
} else if (indices[i].type == ND_INDEX_TYPE_ELLIPSIS) {
num_ellipsis++;
if (num_ellipsis > 1) {
raise_exception(SizeT, EXN_INDEX_ERROR, "an index can only have a single ellipsis ('...')", NO_PARAM,
NO_PARAM, NO_PARAM);
}
} else {
__builtin_unreachable();
}
}
debug_assert_eq(SizeT, expected_dst_ndims, dst_ndarray->ndims);
if (src_ndarray->ndims - num_indexed < 0) {
raise_exception(SizeT, EXN_INDEX_ERROR,
"too many indices for array: array is {0}-dimensional, "
"but {1} were indexed",
src_ndarray->ndims, num_indices, NO_PARAM);
}
dst_ndarray->data = src_ndarray->data;
dst_ndarray->itemsize = src_ndarray->itemsize;
// Reference code:
// https://github.com/wadetb/tinynumpy/blob/0d23d22e07062ffab2afa287374c7b366eebdda1/tinynumpy/tinynumpy.py#L652
SizeT src_axis = 0;
SizeT dst_axis = 0;
for (int32_t i = 0; i < num_indices; i++) {
const NDIndex* index = &indices[i];
if (index->type == ND_INDEX_TYPE_SINGLE_ELEMENT) {
SizeT input = (SizeT) * ((int32_t*)index->data);
SizeT k = slice::resolve_index_in_length(src_ndarray->shape[src_axis], input);
if (k == -1) {
raise_exception(SizeT, EXN_INDEX_ERROR,
"index {0} is out of bounds for axis {1} "
"with size {2}",
input, src_axis, src_ndarray->shape[src_axis]);
}
dst_ndarray->data = static_cast<uint8_t*>(dst_ndarray->data) + k * src_ndarray->strides[src_axis];
src_axis++;
} else if (index->type == ND_INDEX_TYPE_SLICE) {
Slice<int32_t>* slice = (Slice<int32_t>*)index->data;
Range<int32_t> range = slice->indices_checked<SizeT>(src_ndarray->shape[src_axis]);
dst_ndarray->data = static_cast<uint8_t*>(dst_ndarray->data) + (SizeT)range.start * src_ndarray->strides[src_axis];
dst_ndarray->strides[dst_axis] = ((SizeT)range.step) * src_ndarray->strides[src_axis];
dst_ndarray->shape[dst_axis] = (SizeT)range.len<SizeT>();
dst_axis++;
src_axis++;
} else if (index->type == ND_INDEX_TYPE_NEWAXIS) {
dst_ndarray->strides[dst_axis] = 0;
dst_ndarray->shape[dst_axis] = 1;
dst_axis++;
} else if (index->type == ND_INDEX_TYPE_ELLIPSIS) {
// The number of ':' entries this '...' implies.
SizeT ellipsis_size = src_ndarray->ndims - num_indexed;
for (SizeT j = 0; j < ellipsis_size; j++) {
dst_ndarray->strides[dst_axis] = src_ndarray->strides[src_axis];
dst_ndarray->shape[dst_axis] = src_ndarray->shape[src_axis];
dst_axis++;
src_axis++;
}
} else {
__builtin_unreachable();
}
}
for (; dst_axis < dst_ndarray->ndims; dst_axis++, src_axis++) {
dst_ndarray->shape[dst_axis] = src_ndarray->shape[src_axis];
dst_ndarray->strides[dst_axis] = src_ndarray->strides[src_axis];
}
debug_assert_eq(SizeT, src_ndarray->ndims, src_axis);
debug_assert_eq(SizeT, dst_ndarray->ndims, dst_axis);
}
} // namespace indexing
} // namespace ndarray
} // namespace
extern "C" {
using namespace ndarray::indexing;
void __nac3_ndarray_index(int32_t num_indices,
NDIndex* indices,
NDArray<int32_t>* src_ndarray,
NDArray<int32_t>* dst_ndarray) {
index(num_indices, indices, src_ndarray, dst_ndarray);
}
void __nac3_ndarray_index64(int64_t num_indices,
NDIndex* indices,
NDArray<int64_t>* src_ndarray,
NDArray<int64_t>* dst_ndarray) {
index(num_indices, indices, src_ndarray, dst_ndarray);
}
}

View File

@ -0,0 +1,146 @@
#pragma once
#include "irrt/int_types.hpp"
#include "irrt/ndarray/def.hpp"
namespace {
/**
* @brief Helper struct to enumerate through an ndarray *efficiently*.
*
* Example usage (in pseudo-code):
* ```
* // Suppose my_ndarray has been initialized, with shape [2, 3] and dtype `double`
* NDIter nditer;
* nditer.initialize(my_ndarray);
* while (nditer.has_element()) {
* // This body is run 6 (= my_ndarray.size) times.
*
* // [0, 0] -> [0, 1] -> [0, 2] -> [1, 0] -> [1, 1] -> [1, 2] -> end
* print(nditer.indices);
*
* // 0 -> 1 -> 2 -> 3 -> 4 -> 5
* print(nditer.nth);
*
* // <1st element> -> <2nd element> -> ... -> <6th element> -> end
* print(*((double *) nditer.element))
*
* nditer.next(); // Go to next element.
* }
* ```
*
* Interesting cases:
* - If `my_ndarray.ndims` == 0, there is one iteration.
* - If `my_ndarray.shape` contains zeroes, there are no iterations.
*/
template<typename SizeT>
struct NDIter {
// Information about the ndarray being iterated over.
SizeT ndims;
SizeT* shape;
SizeT* strides;
/**
* @brief The current indices.
*
* Must be allocated by the caller.
*/
SizeT* indices;
/**
* @brief The nth (0-based) index of the current indices.
*
* Initially this is 0.
*/
SizeT nth;
/**
* @brief Pointer to the current element.
*
* Initially this points to first element of the ndarray.
*/
void* element;
/**
* @brief Cache for the product of shape.
*
* Could be 0 if `shape` has 0s in it.
*/
SizeT size;
void initialize(SizeT ndims, SizeT* shape, SizeT* strides, void* element, SizeT* indices) {
this->ndims = ndims;
this->shape = shape;
this->strides = strides;
this->indices = indices;
this->element = element;
// Compute size
this->size = 1;
for (SizeT i = 0; i < ndims; i++) {
this->size *= shape[i];
}
// `indices` starts on all 0s.
for (SizeT axis = 0; axis < ndims; axis++)
indices[axis] = 0;
nth = 0;
}
void initialize_by_ndarray(NDArray<SizeT>* ndarray, SizeT* indices) {
// NOTE: ndarray->data is pointing to the first element, and `NDIter`'s `element` should also point to the first
// element as well.
this->initialize(ndarray->ndims, ndarray->shape, ndarray->strides, ndarray->data, indices);
}
// Is the current iteration valid?
// If true, then `element`, `indices` and `nth` contain details about the current element.
bool has_element() { return nth < size; }
// Go to the next element.
void next() {
for (SizeT i = 0; i < ndims; i++) {
SizeT axis = ndims - i - 1;
indices[axis]++;
if (indices[axis] >= shape[axis]) {
indices[axis] = 0;
// TODO: There is something called backstrides to speedup iteration.
// See https://ajcr.net/stride-guide-part-1/, and
// https://docs.scipy.org/doc/numpy-1.13.0/reference/c-api.types-and-structures.html#c.PyArrayIterObject.PyArrayIterObject.backstrides.
element = static_cast<void*>(reinterpret_cast<uint8_t*>(element) - strides[axis] * (shape[axis] - 1));
} else {
element = static_cast<void*>(reinterpret_cast<uint8_t*>(element) + strides[axis]);
break;
}
}
nth++;
}
};
} // namespace
extern "C" {
void __nac3_nditer_initialize(NDIter<int32_t>* iter, NDArray<int32_t>* ndarray, int32_t* indices) {
iter->initialize_by_ndarray(ndarray, indices);
}
void __nac3_nditer_initialize64(NDIter<int64_t>* iter, NDArray<int64_t>* ndarray, int64_t* indices) {
iter->initialize_by_ndarray(ndarray, indices);
}
bool __nac3_nditer_has_element(NDIter<int32_t>* iter) {
return iter->has_element();
}
bool __nac3_nditer_has_element64(NDIter<int64_t>* iter) {
return iter->has_element();
}
void __nac3_nditer_next(NDIter<int32_t>* iter) {
iter->next();
}
void __nac3_nditer_next64(NDIter<int64_t>* iter) {
iter->next();
}
}

View File

@ -0,0 +1,47 @@
#pragma once
#include "irrt/debug.hpp"
#include "irrt/int_types.hpp"
namespace {
namespace range {
template<typename T>
T len(T start, T stop, T step) {
// Reference:
// https://github.com/python/cpython/blob/9dbd12375561a393eaec4b21ee4ac568a407cdb0/Objects/rangeobject.c#L933
if (step > 0 && start < stop)
return 1 + (stop - 1 - start) / step;
else if (step < 0 && start > stop)
return 1 + (start - 1 - stop) / (-step);
else
return 0;
}
} // namespace range
/**
* @brief A Python range.
*/
template<typename T>
struct Range {
T start;
T stop;
T step;
/**
* @brief Calculate the `len()` of this range.
*/
template<typename SizeT>
T len() {
debug_assert(SizeT, step != 0);
return range::len(start, stop, step);
}
};
} // namespace
extern "C" {
using namespace range;
SliceIndex __nac3_range_slice_len(const SliceIndex start, const SliceIndex end, const SliceIndex step) {
return len(start, end, step);
}
}

View File

@ -1,6 +1,145 @@
#pragma once
#include <irrt/int_types.hpp>
#include "irrt/debug.hpp"
#include "irrt/exception.hpp"
#include "irrt/int_types.hpp"
#include "irrt/math_util.hpp"
#include "irrt/range.hpp"
namespace {
namespace slice {
/**
* @brief Resolve a possibly negative index in a list of a known length.
*
* Returns -1 if the resolved index is out of the list's bounds.
*/
template<typename T>
T resolve_index_in_length(T length, T index) {
T resolved = index < 0 ? length + index : index;
if (0 <= resolved && resolved < length) {
return resolved;
} else {
return -1;
}
}
/**
* @brief Resolve a slice as a range.
*
* This is equivalent to `range(*slice(start, stop, step).indices(length))` in Python.
*/
template<typename T>
void indices(bool start_defined,
T start,
bool stop_defined,
T stop,
bool step_defined,
T step,
T length,
T* range_start,
T* range_stop,
T* range_step) {
// Reference: https://github.com/python/cpython/blob/main/Objects/sliceobject.c#L388
*range_step = step_defined ? step : 1;
bool step_is_negative = *range_step < 0;
T lower, upper;
if (step_is_negative) {
lower = -1;
upper = length - 1;
} else {
lower = 0;
upper = length;
}
if (start_defined) {
*range_start = start < 0 ? max(lower, start + length) : min(upper, start);
} else {
*range_start = step_is_negative ? upper : lower;
}
if (stop_defined) {
*range_stop = stop < 0 ? max(lower, stop + length) : min(upper, stop);
} else {
*range_stop = step_is_negative ? lower : upper;
}
}
} // namespace slice
/**
* @brief A Python-like slice with **unresolved** indices.
*/
template<typename T>
struct Slice {
bool start_defined;
T start;
bool stop_defined;
T stop;
bool step_defined;
T step;
Slice() { this->reset(); }
void reset() {
this->start_defined = false;
this->stop_defined = false;
this->step_defined = false;
}
void set_start(T start) {
this->start_defined = true;
this->start = start;
}
void set_stop(T stop) {
this->stop_defined = true;
this->stop = stop;
}
void set_step(T step) {
this->step_defined = true;
this->step = step;
}
/**
* @brief Resolve this slice as a range.
*
* In Python, this would be `range(*slice(start, stop, step).indices(length))`.
*/
template<typename SizeT>
Range<T> indices(T length) {
// Reference:
// https://github.com/python/cpython/blob/main/Objects/sliceobject.c#L388
debug_assert(SizeT, length >= 0);
Range<T> result;
slice::indices(start_defined, start, stop_defined, stop, step_defined, step, length, &result.start,
&result.stop, &result.step);
return result;
}
/**
* @brief Like `.indices()` but with assertions.
*/
template<typename SizeT>
Range<T> indices_checked(T length) {
// TODO: Switch to `SizeT length`
if (length < 0) {
raise_exception(SizeT, EXN_VALUE_ERROR, "length should not be negative, got {0}", length, NO_PARAM,
NO_PARAM);
}
if (this->step_defined && this->step == 0) {
raise_exception(SizeT, EXN_VALUE_ERROR, "slice step cannot be zero", NO_PARAM, NO_PARAM, NO_PARAM);
}
return this->indices<SizeT>(length);
}
};
} // namespace
extern "C" {
SliceIndex __nac3_slice_index_bound(SliceIndex i, const SliceIndex len) {
@ -14,15 +153,4 @@ SliceIndex __nac3_slice_index_bound(SliceIndex i, const SliceIndex len) {
}
return i;
}
SliceIndex __nac3_range_slice_len(const SliceIndex start, const SliceIndex end, const SliceIndex step) {
SliceIndex diff = end - start;
if (diff > 0 && step > 0) {
return ((diff - 1) / step) + 1;
} else if (diff < 0 && step < 0) {
return ((diff + 1) / step) + 1;
} else {
return 0;
}
}
}

View File

@ -0,0 +1,21 @@
[package]
name = "nac3core_derive"
version = "0.1.0"
edition = "2021"
[lib]
proc-macro = true
[[test]]
name = "structfields_tests"
path = "tests/structfields_test.rs"
[dev-dependencies]
nac3core = { path = ".." }
trybuild = { version = "1.0", features = ["diff"] }
[dependencies]
proc-macro2 = "1.0"
proc-macro-error = "1.0"
syn = "2.0"
quote = "1.0"

View File

@ -0,0 +1,320 @@
use proc_macro::TokenStream;
use proc_macro_error::{abort, proc_macro_error};
use quote::quote;
use syn::{
parse_macro_input, spanned::Spanned, Data, DataStruct, Expr, ExprField, ExprMethodCall,
ExprPath, GenericArgument, Ident, LitStr, Path, PathArguments, Type, TypePath,
};
/// Extracts all generic arguments of a [`Type`] into a [`Vec`].
///
/// Returns [`Some`] of a possibly-empty [`Vec`] if the path of `ty` matches with
/// `expected_ty_name`, otherwise returns [`None`].
fn extract_generic_args(expected_ty_name: &'static str, ty: &Type) -> Option<Vec<GenericArgument>> {
let Type::Path(TypePath { qself: None, path, .. }) = ty else {
return None;
};
let segments = &path.segments;
if segments.len() != 1 {
return None;
};
let segment = segments.iter().next().unwrap();
if segment.ident != expected_ty_name {
return None;
}
let PathArguments::AngleBracketed(path_args) = &segment.arguments else {
return Some(Vec::new());
};
let args = &path_args.args;
Some(args.iter().cloned().collect::<Vec<_>>())
}
/// Maps a `path` matching one of the `target_idents` into the `replacement` [`Ident`].
fn map_path_to_ident(path: &Path, target_idents: &[&str], replacement: &str) -> Option<Ident> {
path.require_ident()
.ok()
.filter(|ident| target_idents.iter().any(|target| ident == target))
.map(|ident| Ident::new(replacement, ident.span()))
}
/// Extracts the left-hand side of a dot-expression.
fn extract_dot_operand(expr: &Expr) -> Option<&Expr> {
match expr {
Expr::MethodCall(ExprMethodCall { receiver: operand, .. })
| Expr::Field(ExprField { base: operand, .. }) => Some(operand),
_ => None,
}
}
/// Replaces the top-level receiver of a dot-expression with an [`Ident`], returning `Some(&mut expr)` if the
/// replacement is performed.
///
/// The top-level receiver is the left-most receiver expression, e.g. the top-level receiver of `a.b.c.foo()` is `a`.
fn replace_top_level_receiver(expr: &mut Expr, ident: Ident) -> Option<&mut Expr> {
if let Expr::MethodCall(ExprMethodCall { receiver: operand, .. })
| Expr::Field(ExprField { base: operand, .. }) = expr
{
return if extract_dot_operand(operand).is_some() {
if replace_top_level_receiver(operand, ident).is_some() {
Some(expr)
} else {
None
}
} else {
*operand = Box::new(Expr::Path(ExprPath {
attrs: Vec::default(),
qself: None,
path: ident.into(),
}));
Some(expr)
};
}
None
}
/// Iterates all operands to the left-hand side of the `.` of an [expression][`Expr`], i.e. the container operand of all
/// [`Expr::Field`] and the receiver operand of all [`Expr::MethodCall`].
///
/// The iterator will return the operand expressions in reverse order of appearance. For example, `a.b.c.func()` will
/// return `vec![c, b, a]`.
fn iter_dot_operands(expr: &Expr) -> impl Iterator<Item = &Expr> {
let mut o = extract_dot_operand(expr);
std::iter::from_fn(move || {
let this = o;
o = o.as_ref().and_then(|o| extract_dot_operand(o));
this
})
}
/// Normalizes a value expression for use when creating an instance of this structure, returning a
/// [`proc_macro2::TokenStream`] of tokens representing the normalized expression.
fn normalize_value_expr(expr: &Expr) -> proc_macro2::TokenStream {
match &expr {
Expr::Path(ExprPath { qself: None, path, .. }) => {
if let Some(ident) = map_path_to_ident(path, &["usize", "size_t"], "llvm_usize") {
quote! { #ident }
} else {
abort!(
path,
format!(
"Expected one of `size_t`, `usize`, or an implicit call expression in #[value_type(...)], found {}",
quote!(#expr).to_string(),
)
)
}
}
Expr::Call(_) => {
quote! { ctx.#expr }
}
Expr::MethodCall(_) => {
let base_receiver = iter_dot_operands(expr).last();
match base_receiver {
// `usize.{...}`, `size_t.{...}` -> Rewrite the identifiers to `llvm_usize`
Some(Expr::Path(ExprPath { qself: None, path, .. }))
if map_path_to_ident(path, &["usize", "size_t"], "llvm_usize").is_some() =>
{
let ident =
map_path_to_ident(path, &["usize", "size_t"], "llvm_usize").unwrap();
let mut expr = expr.clone();
let expr = replace_top_level_receiver(&mut expr, ident).unwrap();
quote!(#expr)
}
// `ctx.{...}`, `context.{...}` -> Rewrite the identifiers to `ctx`
Some(Expr::Path(ExprPath { qself: None, path, .. }))
if map_path_to_ident(path, &["ctx", "context"], "ctx").is_some() =>
{
let ident = map_path_to_ident(path, &["ctx", "context"], "ctx").unwrap();
let mut expr = expr.clone();
let expr = replace_top_level_receiver(&mut expr, ident).unwrap();
quote!(#expr)
}
// No reserved identifier prefix -> Prepend `ctx.` to the entire expression
_ => quote! { ctx.#expr },
}
}
_ => {
abort!(
expr,
format!(
"Expected one of `size_t`, `usize`, or an implicit call expression in #[value_type(...)], found {}",
quote!(#expr).to_string(),
)
)
}
}
}
/// Derives an implementation of `codegen::types::structure::StructFields`.
///
/// The benefit of using `#[derive(StructFields)]` is that all index- or order-dependent logic required by
/// `impl StructFields` is automatically generated by this implementation, including the field index as required by
/// `StructField::new` and the fields as returned by `StructFields::to_vec`.
///
/// # Prerequisites
///
/// In order to derive from [`StructFields`], you must implement (or derive) [`Eq`] and [`Copy`] as required by
/// `StructFields`.
///
/// Moreover, `#[derive(StructFields)]` can only be used for `struct`s with named fields, and may only contain fields
/// with either `StructField` or [`PhantomData`] types.
///
/// # Attributes for [`StructFields`]
///
/// Each `StructField` field must be declared with the `#[value_type(...)]` attribute. The argument of `value_type`
/// accepts one of the following:
///
/// - An expression returning an instance of `inkwell::types::BasicType` (with or without the receiver `ctx`/`context`).
/// For example, `context.i8_type()`, `ctx.i8_type()`, and `i8_type()` all refer to `i8`.
/// - The reserved identifiers `usize` and `size_t` referring to an `inkwell::types::IntType` of the platform-dependent
/// integer size. `usize` and `size_t` can also be used as the receiver to other method calls, e.g.
/// `usize.array_type(3)`.
///
/// # Example
///
/// The following is an example of an LLVM slice implemented using `#[derive(StructFields)]`.
///
/// ```rust,ignore
/// use nac3core::{
/// codegen::types::structure::StructField,
/// inkwell::{
/// values::{IntValue, PointerValue},
/// AddressSpace,
/// },
/// };
/// use nac3core_derive::StructFields;
///
/// // All classes that implement StructFields must also implement Eq and Copy
/// #[derive(PartialEq, Eq, Clone, Copy, StructFields)]
/// pub struct SliceValue<'ctx> {
/// // Declares ptr have a value type of i8*
/// //
/// // Can also be written as `ctx.i8_type().ptr_type(...)` or `context.i8_type().ptr_type(...)`
/// #[value_type(i8_type().ptr_type(AddressSpace::default()))]
/// ptr: StructField<'ctx, PointerValue<'ctx>>,
///
/// // Declares len have a value type of usize, depending on the target compilation platform
/// #[value_type(usize)]
/// len: StructField<'ctx, IntValue<'ctx>>,
/// }
/// ```
#[proc_macro_derive(StructFields, attributes(value_type))]
#[proc_macro_error]
pub fn derive(input: TokenStream) -> TokenStream {
let input = parse_macro_input!(input as syn::DeriveInput);
let ident = &input.ident;
let Data::Struct(DataStruct { fields, .. }) = &input.data else {
abort!(input, "Only structs with named fields are supported");
};
if let Err(err_span) =
fields
.iter()
.try_for_each(|field| if field.ident.is_some() { Ok(()) } else { Err(field.span()) })
{
abort!(err_span, "Only structs with named fields are supported");
};
// Check if struct<'ctx>
if input.generics.params.len() != 1 {
abort!(input.generics, "Expected exactly 1 generic parameter")
}
let phantom_info = fields
.iter()
.filter(|field| extract_generic_args("PhantomData", &field.ty).is_some())
.map(|field| field.ident.as_ref().unwrap())
.cloned()
.collect::<Vec<_>>();
let field_info = fields
.iter()
.filter(|field| extract_generic_args("PhantomData", &field.ty).is_none())
.map(|field| {
let ident = field.ident.as_ref().unwrap();
let ty = &field.ty;
let Some(_) = extract_generic_args("StructField", ty) else {
abort!(field, "Only StructField and PhantomData are allowed")
};
let attrs = &field.attrs;
let Some(value_type_attr) =
attrs.iter().find(|attr| attr.path().is_ident("value_type"))
else {
abort!(field, "Expected #[value_type(...)] attribute for field");
};
let Ok(value_type_expr) = value_type_attr.parse_args::<Expr>() else {
abort!(value_type_attr, "Expected expression in #[value_type(...)]");
};
let value_expr_toks = normalize_value_expr(&value_type_expr);
(ident.clone(), value_expr_toks)
})
.collect::<Vec<_>>();
// `<*>::new` impl of `StructField` and `PhantomData` for `StructFields::new`
let phantoms_create = phantom_info
.iter()
.map(|id| quote! { #id: ::std::marker::PhantomData })
.collect::<Vec<_>>();
let fields_create = field_info
.iter()
.map(|(id, ty)| {
let id_lit = LitStr::new(&id.to_string(), id.span());
quote! {
#id: ::nac3core::codegen::types::structure::StructField::create(
&mut counter,
#id_lit,
#ty,
)
}
})
.collect::<Vec<_>>();
// `.into()` impl of `StructField` for `StructFields::to_vec`
let fields_into =
field_info.iter().map(|(id, _)| quote! { self.#id.into() }).collect::<Vec<_>>();
let impl_block = quote! {
impl<'ctx> ::nac3core::codegen::types::structure::StructFields<'ctx> for #ident<'ctx> {
fn new(ctx: impl ::nac3core::inkwell::context::AsContextRef<'ctx>, llvm_usize: ::nac3core::inkwell::types::IntType<'ctx>) -> Self {
let ctx = unsafe { ::nac3core::inkwell::context::ContextRef::new(ctx.as_ctx_ref()) };
let mut counter = ::nac3core::codegen::types::structure::FieldIndexCounter::default();
#ident {
#(#fields_create),*
#(#phantoms_create),*
}
}
fn to_vec(&self) -> ::std::vec::Vec<(&'static str, ::nac3core::inkwell::types::BasicTypeEnum<'ctx>)> {
vec![
#(#fields_into),*
]
}
}
};
impl_block.into()
}

View File

@ -0,0 +1,9 @@
use nac3core_derive::StructFields;
use std::marker::PhantomData;
#[derive(PartialEq, Eq, Clone, Copy, StructFields)]
pub struct EmptyValue<'ctx> {
_phantom: PhantomData<&'ctx ()>,
}
fn main() {}

View File

@ -0,0 +1,20 @@
use nac3core::{
codegen::types::structure::StructField,
inkwell::{
values::{IntValue, PointerValue},
AddressSpace,
},
};
use nac3core_derive::StructFields;
#[derive(PartialEq, Eq, Clone, Copy, StructFields)]
pub struct NDArrayValue<'ctx> {
#[value_type(usize)]
ndims: StructField<'ctx, IntValue<'ctx>>,
#[value_type(usize.ptr_type(AddressSpace::default()))]
shape: StructField<'ctx, PointerValue<'ctx>>,
#[value_type(i8_type().ptr_type(AddressSpace::default()))]
data: StructField<'ctx, PointerValue<'ctx>>,
}
fn main() {}

View File

@ -0,0 +1,18 @@
use nac3core::{
codegen::types::structure::StructField,
inkwell::{
values::{IntValue, PointerValue},
AddressSpace,
},
};
use nac3core_derive::StructFields;
#[derive(PartialEq, Eq, Clone, Copy, StructFields)]
pub struct SliceValue<'ctx> {
#[value_type(i8_type().ptr_type(AddressSpace::default()))]
ptr: StructField<'ctx, PointerValue<'ctx>>,
#[value_type(usize)]
len: StructField<'ctx, IntValue<'ctx>>,
}
fn main() {}

View File

@ -0,0 +1,18 @@
use nac3core::{
codegen::types::structure::StructField,
inkwell::{
values::{IntValue, PointerValue},
AddressSpace,
},
};
use nac3core_derive::StructFields;
#[derive(PartialEq, Eq, Clone, Copy, StructFields)]
pub struct SliceValue<'ctx> {
#[value_type(context.i8_type().ptr_type(AddressSpace::default()))]
ptr: StructField<'ctx, PointerValue<'ctx>>,
#[value_type(usize)]
len: StructField<'ctx, IntValue<'ctx>>,
}
fn main() {}

View File

@ -0,0 +1,18 @@
use nac3core::{
codegen::types::structure::StructField,
inkwell::{
values::{IntValue, PointerValue},
AddressSpace,
},
};
use nac3core_derive::StructFields;
#[derive(PartialEq, Eq, Clone, Copy, StructFields)]
pub struct SliceValue<'ctx> {
#[value_type(ctx.i8_type().ptr_type(AddressSpace::default()))]
ptr: StructField<'ctx, PointerValue<'ctx>>,
#[value_type(usize)]
len: StructField<'ctx, IntValue<'ctx>>,
}
fn main() {}

View File

@ -0,0 +1,18 @@
use nac3core::{
codegen::types::structure::StructField,
inkwell::{
values::{IntValue, PointerValue},
AddressSpace,
},
};
use nac3core_derive::StructFields;
#[derive(PartialEq, Eq, Clone, Copy, StructFields)]
pub struct SliceValue<'ctx> {
#[value_type(i8_type().ptr_type(AddressSpace::default()))]
ptr: StructField<'ctx, PointerValue<'ctx>>,
#[value_type(size_t)]
len: StructField<'ctx, IntValue<'ctx>>,
}
fn main() {}

View File

@ -0,0 +1,10 @@
#[test]
fn test_parse_empty() {
let t = trybuild::TestCases::new();
t.pass("tests/structfields_empty.rs");
t.pass("tests/structfields_slice.rs");
t.pass("tests/structfields_slice_ctx.rs");
t.pass("tests/structfields_slice_context.rs");
t.pass("tests/structfields_slice_sizet.rs");
t.pass("tests/structfields_ndarray.rs");
}

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff

View File

@ -1,3 +1,9 @@
use std::collections::HashMap;
use indexmap::IndexMap;
use nac3parser::ast::StrRef;
use crate::{
symbol_resolver::SymbolValue,
toplevel::DefinitionId,
@ -9,10 +15,6 @@ use crate::{
},
};
use indexmap::IndexMap;
use nac3parser::ast::StrRef;
use std::collections::HashMap;
pub struct ConcreteTypeStore {
store: Vec<ConcreteTypeEnum>,
}

View File

@ -1,9 +1,24 @@
use crate::{
codegen::{
classes::{
ArrayLikeIndexer, ArrayLikeValue, ListType, ListValue, NDArrayValue, ProxyType,
ProxyValue, RangeValue, TypedArrayLikeAccessor, UntypedArrayLikeAccessor,
},
use std::{
cmp::min,
collections::HashMap,
convert::TryInto,
iter::{once, repeat, repeat_with, zip},
};
use inkwell::{
attributes::{Attribute, AttributeLoc},
types::{AnyType, BasicType, BasicTypeEnum},
values::{BasicValueEnum, CallSiteValue, FunctionValue, IntValue, PointerValue, StructValue},
AddressSpace, IntPredicate, OptimizationLevel,
};
use itertools::{chain, izip, Either, Itertools};
use nac3parser::ast::{
self, Boolop, Cmpop, Comprehension, Constant, Expr, ExprKind, Location, Operator, StrRef,
Unaryop,
};
use super::{
concrete_type::{ConcreteFuncArg, ConcreteTypeEnum, ConcreteTypeStore},
gen_in_range_check, get_llvm_abi_type, get_llvm_type, get_va_count_arg_name,
irrt::*,
@ -17,12 +32,19 @@ use crate::{
gen_for_callback_incrementing, gen_if_callback, gen_if_else_expr_callback, gen_raise,
gen_var,
},
CodeGenContext, CodeGenTask, CodeGenerator,
types::{ndarray::NDArrayType, ListType},
values::{
ndarray::{NDArrayValue, RustNDIndex},
ArrayLikeIndexer, ArrayLikeValue, ListValue, ProxyValue, RangeValue,
TypedArrayLikeAccessor, UntypedArrayLikeAccessor,
},
CodeGenContext, CodeGenTask, CodeGenerator,
};
use crate::{
symbol_resolver::{SymbolValue, ValueEnum},
toplevel::{
helper::PrimDef,
numpy::{make_ndarray_ty, unpack_ndarray_var_tys},
helper::{extract_ndims, PrimDef},
numpy::unpack_ndarray_var_tys,
DefinitionId, TopLevelDef,
},
typecheck::{
@ -30,20 +52,6 @@ use crate::{
typedef::{FunSignature, FuncArg, Type, TypeEnum, TypeVarId, Unifier, VarMap},
},
};
use inkwell::{
attributes::{Attribute, AttributeLoc},
types::{AnyType, BasicType, BasicTypeEnum},
values::{BasicValueEnum, CallSiteValue, FunctionValue, IntValue, PointerValue, StructValue},
AddressSpace, IntPredicate, OptimizationLevel,
};
use itertools::{chain, izip, Either, Itertools};
use nac3parser::ast::{
self, Boolop, Cmpop, Comprehension, Constant, Expr, ExprKind, Location, Operator, StrRef,
Unaryop,
};
use std::cmp::min;
use std::iter::{repeat, repeat_with};
use std::{collections::HashMap, convert::TryInto, iter::once, iter::zip};
pub fn get_subst_key(
unifier: &mut Unifier,
@ -551,7 +559,7 @@ impl<'ctx, 'a> CodeGenContext<'ctx, 'a> {
&& val_ty.get_element_type().is_struct_type()
} =>
{
self.builder.build_bitcast(*val, arg_ty, "call_arg_cast").unwrap()
self.builder.build_bit_cast(*val, arg_ty, "call_arg_cast").unwrap()
}
_ => *val,
})
@ -971,6 +979,7 @@ pub fn gen_call<'ctx, G: CodeGenerator>(
TopLevelDef::Class { .. } => {
return Ok(Some(generator.gen_constructor(ctx, fun.0, &def, params)?))
}
TopLevelDef::Variable { .. } => unreachable!(),
}
}
.or_else(|_: String| {
@ -1104,7 +1113,7 @@ pub fn allocate_list<'ctx, G: CodeGenerator + ?Sized>(
// List structure; type { ty*, size_t }
let arr_ty = ListType::new(generator, ctx.ctx, llvm_elem_ty);
let list = arr_ty.new_value(generator, ctx, name);
let list = arr_ty.alloca(generator, ctx, name);
let length = ctx.builder.build_int_z_extend(length, llvm_usize, "").unwrap();
list.store_size(ctx, generator, length);
@ -1160,7 +1169,8 @@ pub fn gen_comprehension<'ctx, G: CodeGenerator>(
TypeEnum::TObj { obj_id, .. }
if *obj_id == ctx.primitives.range.obj_id(&ctx.unifier).unwrap() =>
{
let iter_val = RangeValue::from_ptr_val(iter_val.into_pointer_value(), Some("range"));
let iter_val =
RangeValue::from_pointer_value(iter_val.into_pointer_value(), Some("range"));
let (start, stop, step) = destructure_range(ctx, iter_val);
let diff = ctx.builder.build_int_sub(stop, start, "diff").unwrap();
// add 1 to the length as the value is rounded to zero
@ -1391,8 +1401,10 @@ pub fn gen_binop_expr_with_values<'ctx, G: CodeGenerator>(
let llvm_elem_ty = ctx.get_llvm_type(generator, elem_ty1);
let sizeof_elem = llvm_elem_ty.size_of().unwrap();
let lhs = ListValue::from_ptr_val(left_val.into_pointer_value(), llvm_usize, None);
let rhs = ListValue::from_ptr_val(right_val.into_pointer_value(), llvm_usize, None);
let lhs =
ListValue::from_pointer_value(left_val.into_pointer_value(), llvm_usize, None);
let rhs =
ListValue::from_pointer_value(right_val.into_pointer_value(), llvm_usize, None);
let size = ctx
.builder
@ -1475,7 +1487,7 @@ pub fn gen_binop_expr_with_values<'ctx, G: CodeGenerator>(
codegen_unreachable!(ctx)
};
let list_val =
ListValue::from_ptr_val(list_val.into_pointer_value(), llvm_usize, None);
ListValue::from_pointer_value(list_val.into_pointer_value(), llvm_usize, None);
let int_val = ctx
.builder
.build_int_s_extend(int_val.into_int_value(), llvm_usize, "")
@ -1542,8 +1554,6 @@ pub fn gen_binop_expr_with_values<'ctx, G: CodeGenerator>(
} else if ty1.obj_id(&ctx.unifier).is_some_and(|id| id == PrimDef::NDArray.id())
|| ty2.obj_id(&ctx.unifier).is_some_and(|id| id == PrimDef::NDArray.id())
{
let llvm_usize = generator.get_size_type(ctx.ctx);
let is_ndarray1 = ty1.obj_id(&ctx.unifier).is_some_and(|id| id == PrimDef::NDArray.id());
let is_ndarray2 = ty2.obj_id(&ctx.unifier).is_some_and(|id| id == PrimDef::NDArray.id());
@ -1553,10 +1563,10 @@ pub fn gen_binop_expr_with_values<'ctx, G: CodeGenerator>(
assert!(ctx.unifier.unioned(ndarray_dtype1, ndarray_dtype2));
let left_val =
NDArrayValue::from_ptr_val(left_val.into_pointer_value(), llvm_usize, None);
let right_val =
NDArrayValue::from_ptr_val(right_val.into_pointer_value(), llvm_usize, None);
let left_val = NDArrayType::from_unifier_type(generator, ctx, ty1)
.map_value(left_val.into_pointer_value(), None);
let right_val = NDArrayType::from_unifier_type(generator, ctx, ty2)
.map_value(right_val.into_pointer_value(), None);
let res = if op.base == Operator::MatMult {
// MatMult is the only binop which is not an elementwise op
@ -1580,8 +1590,8 @@ pub fn gen_binop_expr_with_values<'ctx, G: CodeGenerator>(
BinopVariant::Normal => None,
BinopVariant::AugAssign => Some(left_val),
},
(left_val.as_base_value().into(), false),
(right_val.as_base_value().into(), false),
(ty1, left_val.as_base_value().into(), false),
(ty2, right_val.as_base_value().into(), false),
|generator, ctx, (lhs, rhs)| {
gen_binop_expr_with_values(
generator,
@ -1605,9 +1615,10 @@ pub fn gen_binop_expr_with_values<'ctx, G: CodeGenerator>(
} else {
let (ndarray_dtype, _) =
unpack_ndarray_var_tys(&mut ctx.unifier, if is_ndarray1 { ty1 } else { ty2 });
let ndarray_val = NDArrayValue::from_ptr_val(
let ndarray_val =
NDArrayType::from_unifier_type(generator, ctx, if is_ndarray1 { ty1 } else { ty2 })
.map_value(
if is_ndarray1 { left_val } else { right_val }.into_pointer_value(),
llvm_usize,
None,
);
let res = numpy::ndarray_elementwise_binop_impl(
@ -1618,8 +1629,8 @@ pub fn gen_binop_expr_with_values<'ctx, G: CodeGenerator>(
BinopVariant::Normal => None,
BinopVariant::AugAssign => Some(ndarray_val),
},
(left_val, !is_ndarray1),
(right_val, !is_ndarray2),
(ty1, left_val, !is_ndarray1),
(ty2, right_val, !is_ndarray2),
|generator, ctx, (lhs, rhs)| {
gen_binop_expr_with_values(
generator,
@ -1770,7 +1781,12 @@ pub fn gen_unaryop_expr_with_values<'ctx, G: CodeGenerator>(
ast::Unaryop::Invert => ctx.builder.build_not(val, "not").map(Into::into).unwrap(),
ast::Unaryop::Not => ctx
.builder
.build_xor(val, val.get_type().const_all_ones(), "not")
.build_int_compare(
inkwell::IntPredicate::EQ,
val,
val.get_type().const_zero(),
"not",
)
.map(Into::into)
.unwrap(),
ast::Unaryop::UAdd => val.into(),
@ -1792,10 +1808,10 @@ pub fn gen_unaryop_expr_with_values<'ctx, G: CodeGenerator>(
_ => val.into(),
}
} else if ty.obj_id(&ctx.unifier).is_some_and(|id| id == PrimDef::NDArray.id()) {
let llvm_usize = generator.get_size_type(ctx.ctx);
let llvm_ndarray_ty = NDArrayType::from_unifier_type(generator, ctx, ty);
let (ndarray_dtype, _) = unpack_ndarray_var_tys(&mut ctx.unifier, ty);
let val = NDArrayValue::from_ptr_val(val.into_pointer_value(), llvm_usize, None);
let val = llvm_ndarray_ty.map_value(val.into_pointer_value(), None);
// ndarray uses `~` rather than `not` to perform elementwise inversion, convert it before
// passing it to the elementwise codegen function
@ -1869,8 +1885,6 @@ pub fn gen_cmpop_expr_with_values<'ctx, G: CodeGenerator>(
if left_ty.obj_id(&ctx.unifier).is_some_and(|id| id == PrimDef::NDArray.id())
|| right_ty.obj_id(&ctx.unifier).is_some_and(|id| id == PrimDef::NDArray.id())
{
let llvm_usize = generator.get_size_type(ctx.ctx);
let (Some(left_ty), lhs) = left else { codegen_unreachable!(ctx) };
let (Some(right_ty), rhs) = comparators[0] else { codegen_unreachable!(ctx) };
let op = ops[0];
@ -1886,15 +1900,15 @@ pub fn gen_cmpop_expr_with_values<'ctx, G: CodeGenerator>(
assert!(ctx.unifier.unioned(ndarray_dtype1, ndarray_dtype2));
let left_val =
NDArrayValue::from_ptr_val(lhs.into_pointer_value(), llvm_usize, None);
let left_val = NDArrayType::from_unifier_type(generator, ctx, left_ty)
.map_value(lhs.into_pointer_value(), None);
let res = numpy::ndarray_elementwise_binop_impl(
generator,
ctx,
ctx.primitives.bool,
None,
(left_val.as_base_value().into(), false),
(rhs, false),
(left_ty, left_val.as_base_value().into(), false),
(right_ty, rhs, false),
|generator, ctx, (lhs, rhs)| {
let val = gen_cmpop_expr_with_values(
generator,
@ -1925,8 +1939,8 @@ pub fn gen_cmpop_expr_with_values<'ctx, G: CodeGenerator>(
ctx,
ctx.primitives.bool,
None,
(lhs, !is_ndarray1),
(rhs, !is_ndarray2),
(left_ty, lhs, !is_ndarray1),
(right_ty, rhs, !is_ndarray2),
|generator, ctx, (lhs, rhs)| {
let val = gen_cmpop_expr_with_values(
generator,
@ -2189,9 +2203,9 @@ pub fn gen_cmpop_expr_with_values<'ctx, G: CodeGenerator>(
}
let left_val =
ListValue::from_ptr_val(lhs.into_pointer_value(), llvm_usize, None);
ListValue::from_pointer_value(lhs.into_pointer_value(), llvm_usize, None);
let right_val =
ListValue::from_ptr_val(rhs.into_pointer_value(), llvm_usize, None);
ListValue::from_pointer_value(rhs.into_pointer_value(), llvm_usize, None);
Ok(gen_if_else_expr_callback(
generator,
@ -2385,7 +2399,10 @@ pub fn gen_cmpop_expr_with_values<'ctx, G: CodeGenerator>(
})
.map(BasicValueEnum::into_int_value)?;
Ok(ctx.builder.build_not(cmp, "").unwrap())
Ok(ctx.builder.build_not(
generator.bool_to_i1(ctx, cmp),
"",
).unwrap())
},
|_, ctx| {
let bb = ctx.builder.get_insert_block().unwrap();
@ -2505,7 +2522,7 @@ fn gen_ndarray_subscript_expr<'ctx, G: CodeGenerator>(
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
ty: Type,
ndims: Type,
ndims_ty: Type,
v: NDArrayValue<'ctx>,
slice: &Expr<Option<Type>>,
) -> Result<Option<ValueEnum<'ctx>>, String> {
@ -2513,7 +2530,7 @@ fn gen_ndarray_subscript_expr<'ctx, G: CodeGenerator>(
let llvm_i32 = ctx.ctx.i32_type();
let llvm_usize = generator.get_size_type(ctx.ctx);
let TypeEnum::TLiteral { values, .. } = &*ctx.unifier.get_ty_immutable(ndims) else {
let TypeEnum::TLiteral { values, .. } = &*ctx.unifier.get_ty_immutable(ndims_ty) else {
codegen_unreachable!(ctx)
};
@ -2546,14 +2563,6 @@ fn gen_ndarray_subscript_expr<'ctx, G: CodeGenerator>(
_ => 1,
};
let ndarray_ndims_ty = ctx.unifier.get_fresh_literal(
ndims.iter().map(|v| SymbolValue::U64(v - subscripted_dims)).collect(),
None,
);
let ndarray_ty =
make_ndarray_ty(&mut ctx.unifier, &ctx.primitives, Some(ty), Some(ndarray_ndims_ty));
let llvm_pndarray_t = ctx.get_llvm_type(generator, ndarray_ty).into_pointer_type();
let llvm_ndarray_t = llvm_pndarray_t.get_element_type().into_struct_type();
let llvm_ndarray_data_t = ctx.get_llvm_type(generator, ty).as_basic_type_enum();
let sizeof_elem = llvm_ndarray_data_t.size_of().unwrap();
@ -2587,7 +2596,7 @@ fn gen_ndarray_subscript_expr<'ctx, G: CodeGenerator>(
let llvm_i32 = ctx.ctx.i32_type();
let len = unsafe {
v.dim_sizes().get_typed_unchecked(
v.shape().get_typed_unchecked(
ctx,
generator,
&llvm_usize.const_int(dim, true),
@ -2628,7 +2637,7 @@ fn gen_ndarray_subscript_expr<'ctx, G: CodeGenerator>(
ExprKind::Slice { lower, upper, step } => {
let dim_sz = unsafe {
v.dim_sizes().get_typed_unchecked(
v.shape().get_typed_unchecked(
ctx,
generator,
&llvm_usize.const_int(dim, false),
@ -2745,26 +2754,15 @@ fn gen_ndarray_subscript_expr<'ctx, G: CodeGenerator>(
_ => {
// Accessing an element from a multi-dimensional `ndarray`
let Some(index_addr) = make_indices_arr(generator, ctx)? else { return Ok(None) };
let num_dims = extract_ndims(&ctx.unifier, ndims_ty) - 1;
// Create a new array, remove the top dimension from the dimension-size-list, and copy the
// elements over
let subscripted_ndarray =
generator.gen_var_alloc(ctx, llvm_ndarray_t.into(), None)?;
let ndarray = NDArrayValue::from_ptr_val(subscripted_ndarray, llvm_usize, None);
let num_dims = v.load_ndims(ctx);
ndarray.store_ndims(
ctx,
generator,
ctx.builder
.build_int_sub(num_dims, llvm_usize.const_int(1, false), "")
.unwrap(),
);
let ndarray_num_dims = ndarray.load_ndims(ctx);
ndarray.create_dim_sizes(ctx, llvm_usize, ndarray_num_dims);
let ndarray =
NDArrayType::new(generator, ctx.ctx, llvm_ndarray_data_t, Some(num_dims))
.construct_uninitialized(generator, ctx, None);
let ndarray_num_dims = ctx
.builder
@ -2775,7 +2773,7 @@ fn gen_ndarray_subscript_expr<'ctx, G: CodeGenerator>(
)
.unwrap();
let v_dims_src_ptr = unsafe {
v.dim_sizes().ptr_offset_unchecked(
v.shape().ptr_offset_unchecked(
ctx,
generator,
&llvm_usize.const_int(1, false),
@ -2784,7 +2782,7 @@ fn gen_ndarray_subscript_expr<'ctx, G: CodeGenerator>(
};
call_memcpy_generic(
ctx,
ndarray.dim_sizes().base_ptr(ctx, generator),
ndarray.shape().base_ptr(ctx, generator),
v_dims_src_ptr,
ctx.builder
.build_int_mul(ndarray_num_dims, llvm_usize.size_of(), "")
@ -2793,17 +2791,17 @@ fn gen_ndarray_subscript_expr<'ctx, G: CodeGenerator>(
llvm_i1.const_zero(),
);
let ndarray_num_elems = call_ndarray_calc_size(
let ndarray_num_elems = ndarray::call_ndarray_calc_size(
generator,
ctx,
&ndarray.dim_sizes().as_slice_value(ctx, generator),
&ndarray.shape().as_slice_value(ctx, generator),
(None, None),
);
let ndarray_num_elems = ctx
.builder
.build_int_z_extend_or_bit_cast(ndarray_num_elems, sizeof_elem.get_type(), "")
.unwrap();
ndarray.create_data(ctx, llvm_ndarray_data_t, ndarray_num_elems);
unsafe { ndarray.create_data(generator, ctx) };
let v_data_src_ptr = v.data().ptr_offset(ctx, generator, &index_addr, None);
call_memcpy_generic(
@ -2876,7 +2874,31 @@ pub fn gen_expr<'ctx, G: CodeGenerator>(
Some((_, Some(static_value), _)) => ValueEnum::Static(static_value.clone()),
None => {
let resolver = ctx.resolver.clone();
resolver.get_symbol_value(*id, ctx).unwrap()
let value = resolver.get_symbol_value(*id, ctx, generator).unwrap();
let globals = ctx
.top_level
.definitions
.read()
.iter()
.filter_map(|def| {
if let TopLevelDef::Variable { simple_name, ty, .. } = &*def.read() {
Some((*simple_name, *ty))
} else {
None
}
})
.collect_vec();
if let Some((_, ty)) = globals.iter().find(|(name, _)| name == id) {
let ptr = value
.to_basic_value_enum(ctx, generator, *ty)
.map(BasicValueEnum::into_pointer_value)?;
ctx.builder.build_load(ptr, id.to_string().as_str()).map(Into::into).unwrap()
} else {
value
}
}
},
ExprKind::List { elts, .. } => {
@ -3069,48 +3091,53 @@ pub fn gen_expr<'ctx, G: CodeGenerator>(
};
let left = generator.bool_to_i1(ctx, left);
let current = ctx.builder.get_insert_block().unwrap().get_parent().unwrap();
let a_bb = ctx.ctx.append_basic_block(current, "a");
let b_bb = ctx.ctx.append_basic_block(current, "b");
let a_begin_bb = ctx.ctx.append_basic_block(current, "a_begin");
let a_end_bb = ctx.ctx.append_basic_block(current, "a_end");
let b_begin_bb = ctx.ctx.append_basic_block(current, "b_begin");
let b_end_bb = ctx.ctx.append_basic_block(current, "b_end");
let cont_bb = ctx.ctx.append_basic_block(current, "cont");
ctx.builder.build_conditional_branch(left, a_bb, b_bb).unwrap();
ctx.builder.build_conditional_branch(left, a_begin_bb, b_begin_bb).unwrap();
ctx.builder.position_at_end(a_end_bb);
ctx.builder.build_unconditional_branch(cont_bb).unwrap();
ctx.builder.position_at_end(b_end_bb);
ctx.builder.build_unconditional_branch(cont_bb).unwrap();
let (a, b) = match op {
Boolop::Or => {
ctx.builder.position_at_end(a_bb);
ctx.builder.position_at_end(a_begin_bb);
let a = ctx.ctx.i8_type().const_int(1, false);
ctx.builder.build_unconditional_branch(cont_bb).unwrap();
ctx.builder.build_unconditional_branch(a_end_bb).unwrap();
ctx.builder.position_at_end(b_bb);
ctx.builder.position_at_end(b_begin_bb);
let b = if let Some(v) = generator.gen_expr(ctx, &values[1])? {
let b = v
.to_basic_value_enum(ctx, generator, values[1].custom.unwrap())?
.into_int_value();
let b = generator.bool_to_i8(ctx, b);
ctx.builder.build_unconditional_branch(cont_bb).unwrap();
Some(b)
} else {
None
};
ctx.builder.build_unconditional_branch(b_end_bb).unwrap();
(Some(a), b)
}
Boolop::And => {
ctx.builder.position_at_end(a_bb);
ctx.builder.position_at_end(a_begin_bb);
let a = if let Some(v) = generator.gen_expr(ctx, &values[1])? {
let a = v
.to_basic_value_enum(ctx, generator, values[1].custom.unwrap())?
.into_int_value();
let a = generator.bool_to_i8(ctx, a);
ctx.builder.build_unconditional_branch(cont_bb).unwrap();
Some(a)
} else {
None
};
ctx.builder.build_unconditional_branch(a_end_bb).unwrap();
ctx.builder.position_at_end(b_bb);
ctx.builder.position_at_end(b_begin_bb);
let b = ctx.ctx.i8_type().const_zero();
ctx.builder.build_unconditional_branch(cont_bb).unwrap();
ctx.builder.build_unconditional_branch(b_end_bb).unwrap();
(a, Some(b))
}
@ -3120,7 +3147,7 @@ pub fn gen_expr<'ctx, G: CodeGenerator>(
match (a, b) {
(Some(a), Some(b)) => {
let phi = ctx.builder.build_phi(ctx.ctx.i8_type(), "").unwrap();
phi.add_incoming(&[(&a, a_bb), (&b, b_bb)]);
phi.add_incoming(&[(&a, a_end_bb), (&b, b_end_bb)]);
phi.as_basic_value().into()
}
(Some(a), None) => a.into(),
@ -3358,7 +3385,7 @@ pub fn gen_expr<'ctx, G: CodeGenerator>(
} else {
return Ok(None);
};
let v = ListValue::from_ptr_val(v, usize, Some("arr"));
let v = ListValue::from_pointer_value(v, usize, Some("arr"));
let ty = ctx.get_llvm_type(generator, *ty);
if let ExprKind::Slice { lower, upper, step } = &slice.node {
let one = int32.const_int(1, false);
@ -3459,18 +3486,22 @@ pub fn gen_expr<'ctx, G: CodeGenerator>(
v.data().get(ctx, generator, &index, None).into()
}
}
TypeEnum::TObj { obj_id, params, .. } if *obj_id == PrimDef::NDArray.id() => {
let (ty, ndims) = params.iter().map(|(_, ty)| ty).collect_tuple().unwrap();
let v = if let Some(v) = generator.gen_expr(ctx, value)? {
v.to_basic_value_enum(ctx, generator, value.custom.unwrap())?
.into_pointer_value()
} else {
TypeEnum::TObj { obj_id, .. } if *obj_id == PrimDef::NDArray.id() => {
let Some(ndarray) = generator.gen_expr(ctx, value)? else {
return Ok(None);
};
let v = NDArrayValue::from_ptr_val(v, usize, None);
return gen_ndarray_subscript_expr(generator, ctx, *ty, *ndims, v, slice);
let ndarray_ty = value.custom.unwrap();
let ndarray = ndarray.to_basic_value_enum(ctx, generator, ndarray_ty)?;
let ndarray = NDArrayType::from_unifier_type(generator, ctx, ndarray_ty)
.map_value(ndarray.into_pointer_value(), None);
let indices = RustNDIndex::from_subscript_expr(generator, ctx, slice)?;
let result = ndarray
.index(generator, ctx, &indices)
.split_unsized(generator, ctx)
.to_basic_value_enum();
return Ok(Some(ValueEnum::Dynamic(result)));
}
TypeEnum::TTuple { .. } => {
let index: u32 =
@ -3519,3 +3550,97 @@ pub fn gen_expr<'ctx, G: CodeGenerator>(
_ => unimplemented!(),
}))
}
/// Creates a function in the current module and inserts a `call` instruction into the LLVM IR.
#[allow(clippy::too_many_arguments)]
pub fn create_fn_and_call<'ctx>(
ctx: &CodeGenContext<'ctx, '_>,
fn_name: &str,
ret_type: Option<BasicTypeEnum<'ctx>>,
params: &[BasicTypeEnum<'ctx>],
args: &[BasicValueEnum<'ctx>],
is_var_args: bool,
call_value_name: Option<&str>,
configure: Option<&dyn Fn(&FunctionValue<'ctx>)>,
) -> Option<BasicValueEnum<'ctx>> {
let intrinsic_fn = ctx.module.get_function(fn_name).unwrap_or_else(|| {
let params = params.iter().copied().map(BasicTypeEnum::into).collect_vec();
let fn_type = if let Some(ret_type) = ret_type {
ret_type.fn_type(params.as_slice(), is_var_args)
} else {
ctx.ctx.void_type().fn_type(params.as_slice(), is_var_args)
};
ctx.module.add_function(fn_name, fn_type, None)
});
if let Some(configure) = configure {
configure(&intrinsic_fn);
}
let args = args.iter().copied().map(BasicValueEnum::into).collect_vec();
ctx.builder
.build_call(intrinsic_fn, args.as_slice(), call_value_name.unwrap_or_default())
.map(CallSiteValue::try_as_basic_value)
.map(Either::left)
.unwrap()
}
/// Creates a function in the current module and inserts a `call` instruction into the LLVM IR.
///
/// This is a wrapper around [`create_fn_and_call`] for non-vararg function. This function allows
/// parameters and arguments to be specified as tuples to better indicate the expected type and
/// actual value of each parameter-argument pair of the call.
pub fn create_and_call_function<'ctx>(
ctx: &CodeGenContext<'ctx, '_>,
fn_name: &str,
ret_type: Option<BasicTypeEnum<'ctx>>,
params: &[(BasicTypeEnum<'ctx>, BasicValueEnum<'ctx>)],
value_name: Option<&str>,
configure: Option<&dyn Fn(&FunctionValue<'ctx>)>,
) -> Option<BasicValueEnum<'ctx>> {
let param_tys = params.iter().map(|(ty, _)| ty).copied().map(BasicTypeEnum::into).collect_vec();
let arg_values =
params.iter().map(|(_, value)| value).copied().map(BasicValueEnum::into).collect_vec();
create_fn_and_call(
ctx,
fn_name,
ret_type,
param_tys.as_slice(),
arg_values.as_slice(),
false,
value_name,
configure,
)
}
/// Creates a function in the current module and inserts a `call` instruction into the LLVM IR.
///
/// This is a wrapper around [`create_fn_and_call`] for non-vararg function. This function allows
/// only arguments to be specified and performs inference for the parameter types of the function
/// using [`BasicValueEnum::get_type`] on the arguments.
///
/// This function is recommended if it is known that all function arguments match the parameter
/// types of the invoked function.
pub fn infer_and_call_function<'ctx>(
ctx: &CodeGenContext<'ctx, '_>,
fn_name: &str,
ret_type: Option<BasicTypeEnum<'ctx>>,
args: &[BasicValueEnum<'ctx>],
value_name: Option<&str>,
configure: Option<&dyn Fn(&FunctionValue<'ctx>)>,
) -> Option<BasicValueEnum<'ctx>> {
let param_tys = args.iter().map(BasicValueEnum::get_type).collect_vec();
create_fn_and_call(
ctx,
fn_name,
ret_type,
param_tys.as_slice(),
args,
false,
value_name,
configure,
)
}

View File

@ -1,8 +1,10 @@
use inkwell::attributes::{Attribute, AttributeLoc};
use inkwell::values::{BasicValueEnum, CallSiteValue, FloatValue, IntValue};
use inkwell::{
attributes::{Attribute, AttributeLoc},
values::{BasicValueEnum, CallSiteValue, FloatValue, IntValue},
};
use itertools::Either;
use crate::codegen::CodeGenContext;
use super::CodeGenContext;
/// Macro to generate extern function
/// Both function return type and function parameter type are `FloatValue`

View File

@ -1,16 +1,18 @@
use crate::{
codegen::{bool_to_i1, bool_to_i8, classes::ArraySliceValue, expr::*, stmt::*, CodeGenContext},
symbol_resolver::ValueEnum,
toplevel::{DefinitionId, TopLevelDef},
typecheck::typedef::{FunSignature, Type},
};
use inkwell::{
context::Context,
types::{BasicTypeEnum, IntType},
values::{BasicValueEnum, IntValue, PointerValue},
};
use nac3parser::ast::{Expr, Stmt, StrRef};
use super::{bool_to_i1, bool_to_i8, expr::*, stmt::*, values::ArraySliceValue, CodeGenContext};
use crate::{
symbol_resolver::ValueEnum,
toplevel::{DefinitionId, TopLevelDef},
typecheck::typedef::{FunSignature, Type},
};
pub trait CodeGenerator {
/// Return the module name for the code generator.
fn get_name(&self) -> &str;

View File

@ -0,0 +1,162 @@
use inkwell::{
types::BasicTypeEnum,
values::{BasicValueEnum, CallSiteValue, IntValue},
AddressSpace, IntPredicate,
};
use itertools::Either;
use super::calculate_len_for_slice_range;
use crate::codegen::{
macros::codegen_unreachable,
values::{ArrayLikeValue, ListValue},
CodeGenContext, CodeGenerator,
};
/// This function handles 'end' **inclusively**.
/// Order of tuples `assign_idx` and `value_idx` is ('start', 'end', 'step').
/// Negative index should be handled before entering this function
pub fn list_slice_assignment<'ctx, G: CodeGenerator + ?Sized>(
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
ty: BasicTypeEnum<'ctx>,
dest_arr: ListValue<'ctx>,
dest_idx: (IntValue<'ctx>, IntValue<'ctx>, IntValue<'ctx>),
src_arr: ListValue<'ctx>,
src_idx: (IntValue<'ctx>, IntValue<'ctx>, IntValue<'ctx>),
) {
let size_ty = generator.get_size_type(ctx.ctx);
let int8_ptr = ctx.ctx.i8_type().ptr_type(AddressSpace::default());
let int32 = ctx.ctx.i32_type();
let (fun_symbol, elem_ptr_type) = ("__nac3_list_slice_assign_var_size", int8_ptr);
let slice_assign_fun = {
let ty_vec = vec![
int32.into(), // dest start idx
int32.into(), // dest end idx
int32.into(), // dest step
elem_ptr_type.into(), // dest arr ptr
int32.into(), // dest arr len
int32.into(), // src start idx
int32.into(), // src end idx
int32.into(), // src step
elem_ptr_type.into(), // src arr ptr
int32.into(), // src arr len
int32.into(), // size
];
ctx.module.get_function(fun_symbol).unwrap_or_else(|| {
let fn_t = int32.fn_type(ty_vec.as_slice(), false);
ctx.module.add_function(fun_symbol, fn_t, None)
})
};
let zero = int32.const_zero();
let one = int32.const_int(1, false);
let dest_arr_ptr = dest_arr.data().base_ptr(ctx, generator);
let dest_arr_ptr =
ctx.builder.build_pointer_cast(dest_arr_ptr, elem_ptr_type, "dest_arr_ptr_cast").unwrap();
let dest_len = dest_arr.load_size(ctx, Some("dest.len"));
let dest_len = ctx.builder.build_int_truncate_or_bit_cast(dest_len, int32, "srclen32").unwrap();
let src_arr_ptr = src_arr.data().base_ptr(ctx, generator);
let src_arr_ptr =
ctx.builder.build_pointer_cast(src_arr_ptr, elem_ptr_type, "src_arr_ptr_cast").unwrap();
let src_len = src_arr.load_size(ctx, Some("src.len"));
let src_len = ctx.builder.build_int_truncate_or_bit_cast(src_len, int32, "srclen32").unwrap();
// index in bound and positive should be done
// assert if dest.step == 1 then len(src) <= len(dest) else len(src) == len(dest), and
// throw exception if not satisfied
let src_end = ctx
.builder
.build_select(
ctx.builder.build_int_compare(IntPredicate::SLT, src_idx.2, zero, "is_neg").unwrap(),
ctx.builder.build_int_sub(src_idx.1, one, "e_min_one").unwrap(),
ctx.builder.build_int_add(src_idx.1, one, "e_add_one").unwrap(),
"final_e",
)
.map(BasicValueEnum::into_int_value)
.unwrap();
let dest_end = ctx
.builder
.build_select(
ctx.builder.build_int_compare(IntPredicate::SLT, dest_idx.2, zero, "is_neg").unwrap(),
ctx.builder.build_int_sub(dest_idx.1, one, "e_min_one").unwrap(),
ctx.builder.build_int_add(dest_idx.1, one, "e_add_one").unwrap(),
"final_e",
)
.map(BasicValueEnum::into_int_value)
.unwrap();
let src_slice_len =
calculate_len_for_slice_range(generator, ctx, src_idx.0, src_end, src_idx.2);
let dest_slice_len =
calculate_len_for_slice_range(generator, ctx, dest_idx.0, dest_end, dest_idx.2);
let src_eq_dest = ctx
.builder
.build_int_compare(IntPredicate::EQ, src_slice_len, dest_slice_len, "slice_src_eq_dest")
.unwrap();
let src_slt_dest = ctx
.builder
.build_int_compare(IntPredicate::SLT, src_slice_len, dest_slice_len, "slice_src_slt_dest")
.unwrap();
let dest_step_eq_one = ctx
.builder
.build_int_compare(
IntPredicate::EQ,
dest_idx.2,
dest_idx.2.get_type().const_int(1, false),
"slice_dest_step_eq_one",
)
.unwrap();
let cond_1 = ctx.builder.build_and(dest_step_eq_one, src_slt_dest, "slice_cond_1").unwrap();
let cond = ctx.builder.build_or(src_eq_dest, cond_1, "slice_cond").unwrap();
ctx.make_assert(
generator,
cond,
"0:ValueError",
"attempt to assign sequence of size {0} to slice of size {1} with step size {2}",
[Some(src_slice_len), Some(dest_slice_len), Some(dest_idx.2)],
ctx.current_loc,
);
let new_len = {
let args = vec![
dest_idx.0.into(), // dest start idx
dest_idx.1.into(), // dest end idx
dest_idx.2.into(), // dest step
dest_arr_ptr.into(), // dest arr ptr
dest_len.into(), // dest arr len
src_idx.0.into(), // src start idx
src_idx.1.into(), // src end idx
src_idx.2.into(), // src step
src_arr_ptr.into(), // src arr ptr
src_len.into(), // src arr len
{
let s = match ty {
BasicTypeEnum::FloatType(t) => t.size_of(),
BasicTypeEnum::IntType(t) => t.size_of(),
BasicTypeEnum::PointerType(t) => t.size_of(),
BasicTypeEnum::StructType(t) => t.size_of().unwrap(),
_ => codegen_unreachable!(ctx),
};
ctx.builder.build_int_truncate_or_bit_cast(s, int32, "size").unwrap()
}
.into(),
];
ctx.builder
.build_call(slice_assign_fun, args.as_slice(), "slice_assign")
.map(CallSiteValue::try_as_basic_value)
.map(|v| v.map_left(BasicValueEnum::into_int_value))
.map(Either::unwrap_left)
.unwrap()
};
// update length
let need_update =
ctx.builder.build_int_compare(IntPredicate::NE, new_len, dest_len, "need_update").unwrap();
let current = ctx.builder.get_insert_block().unwrap().get_parent().unwrap();
let update_bb = ctx.ctx.append_basic_block(current, "update");
let cont_bb = ctx.ctx.append_basic_block(current, "cont");
ctx.builder.build_conditional_branch(need_update, update_bb, cont_bb).unwrap();
ctx.builder.position_at_end(update_bb);
let new_len = ctx.builder.build_int_z_extend_or_bit_cast(new_len, size_ty, "new_len").unwrap();
dest_arr.store_size(ctx, generator, new_len);
ctx.builder.build_unconditional_branch(cont_bb).unwrap();
ctx.builder.position_at_end(cont_bb);
}

View File

@ -0,0 +1,152 @@
use inkwell::{
values::{BasicValueEnum, CallSiteValue, FloatValue, IntValue},
IntPredicate,
};
use itertools::Either;
use crate::codegen::{
macros::codegen_unreachable,
{CodeGenContext, CodeGenerator},
};
// repeated squaring method adapted from GNU Scientific Library:
// https://git.savannah.gnu.org/cgit/gsl.git/tree/sys/pow_int.c
pub fn integer_power<'ctx, G: CodeGenerator + ?Sized>(
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
base: IntValue<'ctx>,
exp: IntValue<'ctx>,
signed: bool,
) -> IntValue<'ctx> {
let symbol = match (base.get_type().get_bit_width(), exp.get_type().get_bit_width(), signed) {
(32, 32, true) => "__nac3_int_exp_int32_t",
(64, 64, true) => "__nac3_int_exp_int64_t",
(32, 32, false) => "__nac3_int_exp_uint32_t",
(64, 64, false) => "__nac3_int_exp_uint64_t",
_ => codegen_unreachable!(ctx),
};
let base_type = base.get_type();
let pow_fun = ctx.module.get_function(symbol).unwrap_or_else(|| {
let fn_type = base_type.fn_type(&[base_type.into(), base_type.into()], false);
ctx.module.add_function(symbol, fn_type, None)
});
// throw exception when exp < 0
let ge_zero = ctx
.builder
.build_int_compare(
IntPredicate::SGE,
exp,
exp.get_type().const_zero(),
"assert_int_pow_ge_0",
)
.unwrap();
ctx.make_assert(
generator,
ge_zero,
"0:ValueError",
"integer power must be positive or zero",
[None, None, None],
ctx.current_loc,
);
ctx.builder
.build_call(pow_fun, &[base.into(), exp.into()], "call_int_pow")
.map(CallSiteValue::try_as_basic_value)
.map(|v| v.map_left(BasicValueEnum::into_int_value))
.map(Either::unwrap_left)
.unwrap()
}
/// Generates a call to `isinf` in IR. Returns an `i1` representing the result.
pub fn call_isinf<'ctx, G: CodeGenerator + ?Sized>(
generator: &mut G,
ctx: &CodeGenContext<'ctx, '_>,
v: FloatValue<'ctx>,
) -> IntValue<'ctx> {
let intrinsic_fn = ctx.module.get_function("__nac3_isinf").unwrap_or_else(|| {
let fn_type = ctx.ctx.i32_type().fn_type(&[ctx.ctx.f64_type().into()], false);
ctx.module.add_function("__nac3_isinf", fn_type, None)
});
let ret = ctx
.builder
.build_call(intrinsic_fn, &[v.into()], "isinf")
.map(CallSiteValue::try_as_basic_value)
.map(|v| v.map_left(BasicValueEnum::into_int_value))
.map(Either::unwrap_left)
.unwrap();
generator.bool_to_i1(ctx, ret)
}
/// Generates a call to `isnan` in IR. Returns an `i1` representing the result.
pub fn call_isnan<'ctx, G: CodeGenerator + ?Sized>(
generator: &mut G,
ctx: &CodeGenContext<'ctx, '_>,
v: FloatValue<'ctx>,
) -> IntValue<'ctx> {
let intrinsic_fn = ctx.module.get_function("__nac3_isnan").unwrap_or_else(|| {
let fn_type = ctx.ctx.i32_type().fn_type(&[ctx.ctx.f64_type().into()], false);
ctx.module.add_function("__nac3_isnan", fn_type, None)
});
let ret = ctx
.builder
.build_call(intrinsic_fn, &[v.into()], "isnan")
.map(CallSiteValue::try_as_basic_value)
.map(|v| v.map_left(BasicValueEnum::into_int_value))
.map(Either::unwrap_left)
.unwrap();
generator.bool_to_i1(ctx, ret)
}
/// Generates a call to `gamma` in IR. Returns an `f64` representing the result.
pub fn call_gamma<'ctx>(ctx: &CodeGenContext<'ctx, '_>, v: FloatValue<'ctx>) -> FloatValue<'ctx> {
let llvm_f64 = ctx.ctx.f64_type();
let intrinsic_fn = ctx.module.get_function("__nac3_gamma").unwrap_or_else(|| {
let fn_type = llvm_f64.fn_type(&[llvm_f64.into()], false);
ctx.module.add_function("__nac3_gamma", fn_type, None)
});
ctx.builder
.build_call(intrinsic_fn, &[v.into()], "gamma")
.map(CallSiteValue::try_as_basic_value)
.map(|v| v.map_left(BasicValueEnum::into_float_value))
.map(Either::unwrap_left)
.unwrap()
}
/// Generates a call to `gammaln` in IR. Returns an `f64` representing the result.
pub fn call_gammaln<'ctx>(ctx: &CodeGenContext<'ctx, '_>, v: FloatValue<'ctx>) -> FloatValue<'ctx> {
let llvm_f64 = ctx.ctx.f64_type();
let intrinsic_fn = ctx.module.get_function("__nac3_gammaln").unwrap_or_else(|| {
let fn_type = llvm_f64.fn_type(&[llvm_f64.into()], false);
ctx.module.add_function("__nac3_gammaln", fn_type, None)
});
ctx.builder
.build_call(intrinsic_fn, &[v.into()], "gammaln")
.map(CallSiteValue::try_as_basic_value)
.map(|v| v.map_left(BasicValueEnum::into_float_value))
.map(Either::unwrap_left)
.unwrap()
}
/// Generates a call to `j0` in IR. Returns an `f64` representing the result.
pub fn call_j0<'ctx>(ctx: &CodeGenContext<'ctx, '_>, v: FloatValue<'ctx>) -> FloatValue<'ctx> {
let llvm_f64 = ctx.ctx.f64_type();
let intrinsic_fn = ctx.module.get_function("__nac3_j0").unwrap_or_else(|| {
let fn_type = llvm_f64.fn_type(&[llvm_f64.into()], false);
ctx.module.add_function("__nac3_j0", fn_type, None)
});
ctx.builder
.build_call(intrinsic_fn, &[v.into()], "j0")
.map(CallSiteValue::try_as_basic_value)
.map(|v| v.map_left(BasicValueEnum::into_float_value))
.map(Either::unwrap_left)
.unwrap()
}

View File

@ -1,27 +1,27 @@
use crate::{symbol_resolver::SymbolResolver, typecheck::typedef::Type};
use super::{
classes::{
ArrayLikeIndexer, ArrayLikeValue, ArraySliceValue, ListValue, NDArrayValue,
TypedArrayLikeAccessor, TypedArrayLikeAdapter, UntypedArrayLikeAccessor,
},
llvm_intrinsics,
macros::codegen_unreachable,
stmt::gen_for_callback_incrementing,
CodeGenContext, CodeGenerator,
};
use inkwell::{
attributes::{Attribute, AttributeLoc},
context::Context,
memory_buffer::MemoryBuffer,
module::Module,
types::{BasicTypeEnum, IntType},
values::{BasicValue, BasicValueEnum, CallSiteValue, FloatValue, IntValue},
AddressSpace, IntPredicate,
values::{BasicValue, BasicValueEnum, IntValue},
IntPredicate,
};
use itertools::Either;
use nac3parser::ast::Expr;
use super::{CodeGenContext, CodeGenerator};
use crate::{symbol_resolver::SymbolResolver, typecheck::typedef::Type};
pub use list::*;
pub use math::*;
pub use range::*;
pub use slice::*;
mod list;
mod math;
pub mod ndarray;
mod range;
mod slice;
#[must_use]
pub fn load_irrt<'ctx>(ctx: &'ctx Context, symbol_resolver: &dyn SymbolResolver) -> Module<'ctx> {
let bitcode_buf = MemoryBuffer::create_from_memory_range(
@ -61,86 +61,25 @@ pub fn load_irrt<'ctx>(ctx: &'ctx Context, symbol_resolver: &dyn SymbolResolver)
irrt_mod
}
// repeated squaring method adapted from GNU Scientific Library:
// https://git.savannah.gnu.org/cgit/gsl.git/tree/sys/pow_int.c
pub fn integer_power<'ctx, G: CodeGenerator + ?Sized>(
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
base: IntValue<'ctx>,
exp: IntValue<'ctx>,
signed: bool,
) -> IntValue<'ctx> {
let symbol = match (base.get_type().get_bit_width(), exp.get_type().get_bit_width(), signed) {
(32, 32, true) => "__nac3_int_exp_int32_t",
(64, 64, true) => "__nac3_int_exp_int64_t",
(32, 32, false) => "__nac3_int_exp_uint32_t",
(64, 64, false) => "__nac3_int_exp_uint64_t",
_ => codegen_unreachable!(ctx),
};
let base_type = base.get_type();
let pow_fun = ctx.module.get_function(symbol).unwrap_or_else(|| {
let fn_type = base_type.fn_type(&[base_type.into(), base_type.into()], false);
ctx.module.add_function(symbol, fn_type, None)
});
// throw exception when exp < 0
let ge_zero = ctx
.builder
.build_int_compare(
IntPredicate::SGE,
exp,
exp.get_type().const_zero(),
"assert_int_pow_ge_0",
)
.unwrap();
ctx.make_assert(
generator,
ge_zero,
"0:ValueError",
"integer power must be positive or zero",
[None, None, None],
ctx.current_loc,
);
ctx.builder
.build_call(pow_fun, &[base.into(), exp.into()], "call_int_pow")
.map(CallSiteValue::try_as_basic_value)
.map(|v| v.map_left(BasicValueEnum::into_int_value))
.map(Either::unwrap_left)
.unwrap()
}
pub fn calculate_len_for_slice_range<'ctx, G: CodeGenerator + ?Sized>(
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
start: IntValue<'ctx>,
end: IntValue<'ctx>,
step: IntValue<'ctx>,
) -> IntValue<'ctx> {
const SYMBOL: &str = "__nac3_range_slice_len";
let len_func = ctx.module.get_function(SYMBOL).unwrap_or_else(|| {
let i32_t = ctx.ctx.i32_type();
let fn_t = i32_t.fn_type(&[i32_t.into(), i32_t.into(), i32_t.into()], false);
ctx.module.add_function(SYMBOL, fn_t, None)
});
// assert step != 0, throw exception if not
let not_zero = ctx
.builder
.build_int_compare(IntPredicate::NE, step, step.get_type().const_zero(), "range_step_ne")
.unwrap();
ctx.make_assert(
generator,
not_zero,
"0:ValueError",
"step must not be zero",
[None, None, None],
ctx.current_loc,
);
ctx.builder
.build_call(len_func, &[start.into(), end.into(), step.into()], "calc_len")
.map(CallSiteValue::try_as_basic_value)
.map(|v| v.map_left(BasicValueEnum::into_int_value))
.map(Either::unwrap_left)
.unwrap()
/// Returns the name of a function which contains variants for 32-bit and 64-bit `size_t`.
///
/// - When [`TypeContext::size_type`] is 32-bits, the function name is `fn_name}`.
/// - When [`TypeContext::size_type`] is 64-bits, the function name is `{fn_name}64`.
#[must_use]
pub fn get_usize_dependent_function_name<G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &CodeGenContext<'_, '_>,
name: &str,
) -> String {
let mut name = name.to_owned();
match generator.get_size_type(ctx.ctx).get_bit_width() {
32 => {}
64 => name.push_str("64"),
bit_width => {
panic!("Unsupported int type bit width {bit_width}, must be either 32-bits or 64-bits")
}
}
name
}
/// NOTE: the output value of the end index of this function should be compared ***inclusively***,
@ -308,644 +247,3 @@ pub fn handle_slice_indices<'ctx, G: CodeGenerator>(
}
}))
}
/// this function allows index out of range, since python
/// allows index out of range in slice (`a = [1,2,3]; a[1:10] == [2,3]`).
pub fn handle_slice_index_bound<'ctx, G: CodeGenerator>(
i: &Expr<Option<Type>>,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
length: IntValue<'ctx>,
) -> Result<Option<IntValue<'ctx>>, String> {
const SYMBOL: &str = "__nac3_slice_index_bound";
let func = ctx.module.get_function(SYMBOL).unwrap_or_else(|| {
let i32_t = ctx.ctx.i32_type();
let fn_t = i32_t.fn_type(&[i32_t.into(), i32_t.into()], false);
ctx.module.add_function(SYMBOL, fn_t, None)
});
let i = if let Some(v) = generator.gen_expr(ctx, i)? {
v.to_basic_value_enum(ctx, generator, i.custom.unwrap())?
} else {
return Ok(None);
};
Ok(Some(
ctx.builder
.build_call(func, &[i.into(), length.into()], "bounded_ind")
.map(CallSiteValue::try_as_basic_value)
.map(|v| v.map_left(BasicValueEnum::into_int_value))
.map(Either::unwrap_left)
.unwrap(),
))
}
/// This function handles 'end' **inclusively**.
/// Order of tuples `assign_idx` and `value_idx` is ('start', 'end', 'step').
/// Negative index should be handled before entering this function
pub fn list_slice_assignment<'ctx, G: CodeGenerator + ?Sized>(
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
ty: BasicTypeEnum<'ctx>,
dest_arr: ListValue<'ctx>,
dest_idx: (IntValue<'ctx>, IntValue<'ctx>, IntValue<'ctx>),
src_arr: ListValue<'ctx>,
src_idx: (IntValue<'ctx>, IntValue<'ctx>, IntValue<'ctx>),
) {
let size_ty = generator.get_size_type(ctx.ctx);
let int8_ptr = ctx.ctx.i8_type().ptr_type(AddressSpace::default());
let int32 = ctx.ctx.i32_type();
let (fun_symbol, elem_ptr_type) = ("__nac3_list_slice_assign_var_size", int8_ptr);
let slice_assign_fun = {
let ty_vec = vec![
int32.into(), // dest start idx
int32.into(), // dest end idx
int32.into(), // dest step
elem_ptr_type.into(), // dest arr ptr
int32.into(), // dest arr len
int32.into(), // src start idx
int32.into(), // src end idx
int32.into(), // src step
elem_ptr_type.into(), // src arr ptr
int32.into(), // src arr len
int32.into(), // size
];
ctx.module.get_function(fun_symbol).unwrap_or_else(|| {
let fn_t = int32.fn_type(ty_vec.as_slice(), false);
ctx.module.add_function(fun_symbol, fn_t, None)
})
};
let zero = int32.const_zero();
let one = int32.const_int(1, false);
let dest_arr_ptr = dest_arr.data().base_ptr(ctx, generator);
let dest_arr_ptr =
ctx.builder.build_pointer_cast(dest_arr_ptr, elem_ptr_type, "dest_arr_ptr_cast").unwrap();
let dest_len = dest_arr.load_size(ctx, Some("dest.len"));
let dest_len = ctx.builder.build_int_truncate_or_bit_cast(dest_len, int32, "srclen32").unwrap();
let src_arr_ptr = src_arr.data().base_ptr(ctx, generator);
let src_arr_ptr =
ctx.builder.build_pointer_cast(src_arr_ptr, elem_ptr_type, "src_arr_ptr_cast").unwrap();
let src_len = src_arr.load_size(ctx, Some("src.len"));
let src_len = ctx.builder.build_int_truncate_or_bit_cast(src_len, int32, "srclen32").unwrap();
// index in bound and positive should be done
// assert if dest.step == 1 then len(src) <= len(dest) else len(src) == len(dest), and
// throw exception if not satisfied
let src_end = ctx
.builder
.build_select(
ctx.builder.build_int_compare(IntPredicate::SLT, src_idx.2, zero, "is_neg").unwrap(),
ctx.builder.build_int_sub(src_idx.1, one, "e_min_one").unwrap(),
ctx.builder.build_int_add(src_idx.1, one, "e_add_one").unwrap(),
"final_e",
)
.map(BasicValueEnum::into_int_value)
.unwrap();
let dest_end = ctx
.builder
.build_select(
ctx.builder.build_int_compare(IntPredicate::SLT, dest_idx.2, zero, "is_neg").unwrap(),
ctx.builder.build_int_sub(dest_idx.1, one, "e_min_one").unwrap(),
ctx.builder.build_int_add(dest_idx.1, one, "e_add_one").unwrap(),
"final_e",
)
.map(BasicValueEnum::into_int_value)
.unwrap();
let src_slice_len =
calculate_len_for_slice_range(generator, ctx, src_idx.0, src_end, src_idx.2);
let dest_slice_len =
calculate_len_for_slice_range(generator, ctx, dest_idx.0, dest_end, dest_idx.2);
let src_eq_dest = ctx
.builder
.build_int_compare(IntPredicate::EQ, src_slice_len, dest_slice_len, "slice_src_eq_dest")
.unwrap();
let src_slt_dest = ctx
.builder
.build_int_compare(IntPredicate::SLT, src_slice_len, dest_slice_len, "slice_src_slt_dest")
.unwrap();
let dest_step_eq_one = ctx
.builder
.build_int_compare(
IntPredicate::EQ,
dest_idx.2,
dest_idx.2.get_type().const_int(1, false),
"slice_dest_step_eq_one",
)
.unwrap();
let cond_1 = ctx.builder.build_and(dest_step_eq_one, src_slt_dest, "slice_cond_1").unwrap();
let cond = ctx.builder.build_or(src_eq_dest, cond_1, "slice_cond").unwrap();
ctx.make_assert(
generator,
cond,
"0:ValueError",
"attempt to assign sequence of size {0} to slice of size {1} with step size {2}",
[Some(src_slice_len), Some(dest_slice_len), Some(dest_idx.2)],
ctx.current_loc,
);
let new_len = {
let args = vec![
dest_idx.0.into(), // dest start idx
dest_idx.1.into(), // dest end idx
dest_idx.2.into(), // dest step
dest_arr_ptr.into(), // dest arr ptr
dest_len.into(), // dest arr len
src_idx.0.into(), // src start idx
src_idx.1.into(), // src end idx
src_idx.2.into(), // src step
src_arr_ptr.into(), // src arr ptr
src_len.into(), // src arr len
{
let s = match ty {
BasicTypeEnum::FloatType(t) => t.size_of(),
BasicTypeEnum::IntType(t) => t.size_of(),
BasicTypeEnum::PointerType(t) => t.size_of(),
BasicTypeEnum::StructType(t) => t.size_of().unwrap(),
_ => codegen_unreachable!(ctx),
};
ctx.builder.build_int_truncate_or_bit_cast(s, int32, "size").unwrap()
}
.into(),
];
ctx.builder
.build_call(slice_assign_fun, args.as_slice(), "slice_assign")
.map(CallSiteValue::try_as_basic_value)
.map(|v| v.map_left(BasicValueEnum::into_int_value))
.map(Either::unwrap_left)
.unwrap()
};
// update length
let need_update =
ctx.builder.build_int_compare(IntPredicate::NE, new_len, dest_len, "need_update").unwrap();
let current = ctx.builder.get_insert_block().unwrap().get_parent().unwrap();
let update_bb = ctx.ctx.append_basic_block(current, "update");
let cont_bb = ctx.ctx.append_basic_block(current, "cont");
ctx.builder.build_conditional_branch(need_update, update_bb, cont_bb).unwrap();
ctx.builder.position_at_end(update_bb);
let new_len = ctx.builder.build_int_z_extend_or_bit_cast(new_len, size_ty, "new_len").unwrap();
dest_arr.store_size(ctx, generator, new_len);
ctx.builder.build_unconditional_branch(cont_bb).unwrap();
ctx.builder.position_at_end(cont_bb);
}
/// Generates a call to `isinf` in IR. Returns an `i1` representing the result.
pub fn call_isinf<'ctx, G: CodeGenerator + ?Sized>(
generator: &mut G,
ctx: &CodeGenContext<'ctx, '_>,
v: FloatValue<'ctx>,
) -> IntValue<'ctx> {
let intrinsic_fn = ctx.module.get_function("__nac3_isinf").unwrap_or_else(|| {
let fn_type = ctx.ctx.i32_type().fn_type(&[ctx.ctx.f64_type().into()], false);
ctx.module.add_function("__nac3_isinf", fn_type, None)
});
let ret = ctx
.builder
.build_call(intrinsic_fn, &[v.into()], "isinf")
.map(CallSiteValue::try_as_basic_value)
.map(|v| v.map_left(BasicValueEnum::into_int_value))
.map(Either::unwrap_left)
.unwrap();
generator.bool_to_i1(ctx, ret)
}
/// Generates a call to `isnan` in IR. Returns an `i1` representing the result.
pub fn call_isnan<'ctx, G: CodeGenerator + ?Sized>(
generator: &mut G,
ctx: &CodeGenContext<'ctx, '_>,
v: FloatValue<'ctx>,
) -> IntValue<'ctx> {
let intrinsic_fn = ctx.module.get_function("__nac3_isnan").unwrap_or_else(|| {
let fn_type = ctx.ctx.i32_type().fn_type(&[ctx.ctx.f64_type().into()], false);
ctx.module.add_function("__nac3_isnan", fn_type, None)
});
let ret = ctx
.builder
.build_call(intrinsic_fn, &[v.into()], "isnan")
.map(CallSiteValue::try_as_basic_value)
.map(|v| v.map_left(BasicValueEnum::into_int_value))
.map(Either::unwrap_left)
.unwrap();
generator.bool_to_i1(ctx, ret)
}
/// Generates a call to `gamma` in IR. Returns an `f64` representing the result.
pub fn call_gamma<'ctx>(ctx: &CodeGenContext<'ctx, '_>, v: FloatValue<'ctx>) -> FloatValue<'ctx> {
let llvm_f64 = ctx.ctx.f64_type();
let intrinsic_fn = ctx.module.get_function("__nac3_gamma").unwrap_or_else(|| {
let fn_type = llvm_f64.fn_type(&[llvm_f64.into()], false);
ctx.module.add_function("__nac3_gamma", fn_type, None)
});
ctx.builder
.build_call(intrinsic_fn, &[v.into()], "gamma")
.map(CallSiteValue::try_as_basic_value)
.map(|v| v.map_left(BasicValueEnum::into_float_value))
.map(Either::unwrap_left)
.unwrap()
}
/// Generates a call to `gammaln` in IR. Returns an `f64` representing the result.
pub fn call_gammaln<'ctx>(ctx: &CodeGenContext<'ctx, '_>, v: FloatValue<'ctx>) -> FloatValue<'ctx> {
let llvm_f64 = ctx.ctx.f64_type();
let intrinsic_fn = ctx.module.get_function("__nac3_gammaln").unwrap_or_else(|| {
let fn_type = llvm_f64.fn_type(&[llvm_f64.into()], false);
ctx.module.add_function("__nac3_gammaln", fn_type, None)
});
ctx.builder
.build_call(intrinsic_fn, &[v.into()], "gammaln")
.map(CallSiteValue::try_as_basic_value)
.map(|v| v.map_left(BasicValueEnum::into_float_value))
.map(Either::unwrap_left)
.unwrap()
}
/// Generates a call to `j0` in IR. Returns an `f64` representing the result.
pub fn call_j0<'ctx>(ctx: &CodeGenContext<'ctx, '_>, v: FloatValue<'ctx>) -> FloatValue<'ctx> {
let llvm_f64 = ctx.ctx.f64_type();
let intrinsic_fn = ctx.module.get_function("__nac3_j0").unwrap_or_else(|| {
let fn_type = llvm_f64.fn_type(&[llvm_f64.into()], false);
ctx.module.add_function("__nac3_j0", fn_type, None)
});
ctx.builder
.build_call(intrinsic_fn, &[v.into()], "j0")
.map(CallSiteValue::try_as_basic_value)
.map(|v| v.map_left(BasicValueEnum::into_float_value))
.map(Either::unwrap_left)
.unwrap()
}
/// Generates a call to `__nac3_ndarray_calc_size`. Returns an [`IntValue`] representing the
/// calculated total size.
///
/// * `dims` - An [`ArrayLikeIndexer`] containing the size of each dimension.
/// * `range` - The dimension index to begin and end (exclusively) calculating the dimensions for,
/// or [`None`] if starting from the first dimension and ending at the last dimension
/// respectively.
pub fn call_ndarray_calc_size<'ctx, G, Dims>(
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
dims: &Dims,
(begin, end): (Option<IntValue<'ctx>>, Option<IntValue<'ctx>>),
) -> IntValue<'ctx>
where
G: CodeGenerator + ?Sized,
Dims: ArrayLikeIndexer<'ctx>,
{
let llvm_usize = generator.get_size_type(ctx.ctx);
let llvm_pusize = llvm_usize.ptr_type(AddressSpace::default());
let ndarray_calc_size_fn_name = match llvm_usize.get_bit_width() {
32 => "__nac3_ndarray_calc_size",
64 => "__nac3_ndarray_calc_size64",
bw => codegen_unreachable!(ctx, "Unsupported size type bit width: {}", bw),
};
let ndarray_calc_size_fn_t = llvm_usize.fn_type(
&[llvm_pusize.into(), llvm_usize.into(), llvm_usize.into(), llvm_usize.into()],
false,
);
let ndarray_calc_size_fn =
ctx.module.get_function(ndarray_calc_size_fn_name).unwrap_or_else(|| {
ctx.module.add_function(ndarray_calc_size_fn_name, ndarray_calc_size_fn_t, None)
});
let begin = begin.unwrap_or_else(|| llvm_usize.const_zero());
let end = end.unwrap_or_else(|| dims.size(ctx, generator));
ctx.builder
.build_call(
ndarray_calc_size_fn,
&[
dims.base_ptr(ctx, generator).into(),
dims.size(ctx, generator).into(),
begin.into(),
end.into(),
],
"",
)
.map(CallSiteValue::try_as_basic_value)
.map(|v| v.map_left(BasicValueEnum::into_int_value))
.map(Either::unwrap_left)
.unwrap()
}
/// Generates a call to `__nac3_ndarray_calc_nd_indices`. Returns a [`TypeArrayLikeAdpater`]
/// containing `i32` indices of the flattened index.
///
/// * `index` - The index to compute the multidimensional index for.
/// * `ndarray` - LLVM pointer to the `NDArray`. This value must be the LLVM representation of an
/// `NDArray`.
pub fn call_ndarray_calc_nd_indices<'ctx, G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &mut CodeGenContext<'ctx, '_>,
index: IntValue<'ctx>,
ndarray: NDArrayValue<'ctx>,
) -> TypedArrayLikeAdapter<'ctx, IntValue<'ctx>> {
let llvm_void = ctx.ctx.void_type();
let llvm_i32 = ctx.ctx.i32_type();
let llvm_usize = generator.get_size_type(ctx.ctx);
let llvm_pi32 = llvm_i32.ptr_type(AddressSpace::default());
let llvm_pusize = llvm_usize.ptr_type(AddressSpace::default());
let ndarray_calc_nd_indices_fn_name = match llvm_usize.get_bit_width() {
32 => "__nac3_ndarray_calc_nd_indices",
64 => "__nac3_ndarray_calc_nd_indices64",
bw => codegen_unreachable!(ctx, "Unsupported size type bit width: {}", bw),
};
let ndarray_calc_nd_indices_fn =
ctx.module.get_function(ndarray_calc_nd_indices_fn_name).unwrap_or_else(|| {
let fn_type = llvm_void.fn_type(
&[llvm_usize.into(), llvm_pusize.into(), llvm_usize.into(), llvm_pi32.into()],
false,
);
ctx.module.add_function(ndarray_calc_nd_indices_fn_name, fn_type, None)
});
let ndarray_num_dims = ndarray.load_ndims(ctx);
let ndarray_dims = ndarray.dim_sizes();
let indices = ctx.builder.build_array_alloca(llvm_i32, ndarray_num_dims, "").unwrap();
ctx.builder
.build_call(
ndarray_calc_nd_indices_fn,
&[
index.into(),
ndarray_dims.base_ptr(ctx, generator).into(),
ndarray_num_dims.into(),
indices.into(),
],
"",
)
.unwrap();
TypedArrayLikeAdapter::from(
ArraySliceValue::from_ptr_val(indices, ndarray_num_dims, None),
Box::new(|_, v| v.into_int_value()),
Box::new(|_, v| v.into()),
)
}
fn call_ndarray_flatten_index_impl<'ctx, G, Indices>(
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
ndarray: NDArrayValue<'ctx>,
indices: &Indices,
) -> IntValue<'ctx>
where
G: CodeGenerator + ?Sized,
Indices: ArrayLikeIndexer<'ctx>,
{
let llvm_i32 = ctx.ctx.i32_type();
let llvm_usize = generator.get_size_type(ctx.ctx);
let llvm_pi32 = llvm_i32.ptr_type(AddressSpace::default());
let llvm_pusize = llvm_usize.ptr_type(AddressSpace::default());
debug_assert_eq!(
IntType::try_from(indices.element_type(ctx, generator))
.map(IntType::get_bit_width)
.unwrap_or_default(),
llvm_i32.get_bit_width(),
"Expected i32 value for argument `indices` to `call_ndarray_flatten_index_impl`"
);
debug_assert_eq!(
indices.size(ctx, generator).get_type().get_bit_width(),
llvm_usize.get_bit_width(),
"Expected usize integer value for argument `indices_size` to `call_ndarray_flatten_index_impl`"
);
let ndarray_flatten_index_fn_name = match llvm_usize.get_bit_width() {
32 => "__nac3_ndarray_flatten_index",
64 => "__nac3_ndarray_flatten_index64",
bw => codegen_unreachable!(ctx, "Unsupported size type bit width: {}", bw),
};
let ndarray_flatten_index_fn =
ctx.module.get_function(ndarray_flatten_index_fn_name).unwrap_or_else(|| {
let fn_type = llvm_usize.fn_type(
&[llvm_pusize.into(), llvm_usize.into(), llvm_pi32.into(), llvm_usize.into()],
false,
);
ctx.module.add_function(ndarray_flatten_index_fn_name, fn_type, None)
});
let ndarray_num_dims = ndarray.load_ndims(ctx);
let ndarray_dims = ndarray.dim_sizes();
let index = ctx
.builder
.build_call(
ndarray_flatten_index_fn,
&[
ndarray_dims.base_ptr(ctx, generator).into(),
ndarray_num_dims.into(),
indices.base_ptr(ctx, generator).into(),
indices.size(ctx, generator).into(),
],
"",
)
.map(CallSiteValue::try_as_basic_value)
.map(|v| v.map_left(BasicValueEnum::into_int_value))
.map(Either::unwrap_left)
.unwrap();
index
}
/// Generates a call to `__nac3_ndarray_flatten_index`. Returns the flattened index for the
/// multidimensional index.
///
/// * `ndarray` - LLVM pointer to the `NDArray`. This value must be the LLVM representation of an
/// `NDArray`.
/// * `indices` - The multidimensional index to compute the flattened index for.
pub fn call_ndarray_flatten_index<'ctx, G, Index>(
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
ndarray: NDArrayValue<'ctx>,
indices: &Index,
) -> IntValue<'ctx>
where
G: CodeGenerator + ?Sized,
Index: ArrayLikeIndexer<'ctx>,
{
call_ndarray_flatten_index_impl(generator, ctx, ndarray, indices)
}
/// Generates a call to `__nac3_ndarray_calc_broadcast`. Returns a tuple containing the number of
/// dimension and size of each dimension of the resultant `ndarray`.
pub fn call_ndarray_calc_broadcast<'ctx, G: CodeGenerator + ?Sized>(
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
lhs: NDArrayValue<'ctx>,
rhs: NDArrayValue<'ctx>,
) -> TypedArrayLikeAdapter<'ctx, IntValue<'ctx>> {
let llvm_usize = generator.get_size_type(ctx.ctx);
let llvm_pusize = llvm_usize.ptr_type(AddressSpace::default());
let ndarray_calc_broadcast_fn_name = match llvm_usize.get_bit_width() {
32 => "__nac3_ndarray_calc_broadcast",
64 => "__nac3_ndarray_calc_broadcast64",
bw => codegen_unreachable!(ctx, "Unsupported size type bit width: {}", bw),
};
let ndarray_calc_broadcast_fn =
ctx.module.get_function(ndarray_calc_broadcast_fn_name).unwrap_or_else(|| {
let fn_type = llvm_usize.fn_type(
&[
llvm_pusize.into(),
llvm_usize.into(),
llvm_pusize.into(),
llvm_usize.into(),
llvm_pusize.into(),
],
false,
);
ctx.module.add_function(ndarray_calc_broadcast_fn_name, fn_type, None)
});
let lhs_ndims = lhs.load_ndims(ctx);
let rhs_ndims = rhs.load_ndims(ctx);
let min_ndims = llvm_intrinsics::call_int_umin(ctx, lhs_ndims, rhs_ndims, None);
gen_for_callback_incrementing(
generator,
ctx,
None,
llvm_usize.const_zero(),
(min_ndims, false),
|generator, ctx, _, idx| {
let idx = ctx.builder.build_int_sub(min_ndims, idx, "").unwrap();
let (lhs_dim_sz, rhs_dim_sz) = unsafe {
(
lhs.dim_sizes().get_typed_unchecked(ctx, generator, &idx, None),
rhs.dim_sizes().get_typed_unchecked(ctx, generator, &idx, None),
)
};
let llvm_usize_const_one = llvm_usize.const_int(1, false);
let lhs_eqz = ctx
.builder
.build_int_compare(IntPredicate::EQ, lhs_dim_sz, llvm_usize_const_one, "")
.unwrap();
let rhs_eqz = ctx
.builder
.build_int_compare(IntPredicate::EQ, rhs_dim_sz, llvm_usize_const_one, "")
.unwrap();
let lhs_or_rhs_eqz = ctx.builder.build_or(lhs_eqz, rhs_eqz, "").unwrap();
let lhs_eq_rhs = ctx
.builder
.build_int_compare(IntPredicate::EQ, lhs_dim_sz, rhs_dim_sz, "")
.unwrap();
let is_compatible = ctx.builder.build_or(lhs_or_rhs_eqz, lhs_eq_rhs, "").unwrap();
ctx.make_assert(
generator,
is_compatible,
"0:ValueError",
"operands could not be broadcast together",
[None, None, None],
ctx.current_loc,
);
Ok(())
},
llvm_usize.const_int(1, false),
)
.unwrap();
let max_ndims = llvm_intrinsics::call_int_umax(ctx, lhs_ndims, rhs_ndims, None);
let lhs_dims = lhs.dim_sizes().base_ptr(ctx, generator);
let lhs_ndims = lhs.load_ndims(ctx);
let rhs_dims = rhs.dim_sizes().base_ptr(ctx, generator);
let rhs_ndims = rhs.load_ndims(ctx);
let out_dims = ctx.builder.build_array_alloca(llvm_usize, max_ndims, "").unwrap();
let out_dims = ArraySliceValue::from_ptr_val(out_dims, max_ndims, None);
ctx.builder
.build_call(
ndarray_calc_broadcast_fn,
&[
lhs_dims.into(),
lhs_ndims.into(),
rhs_dims.into(),
rhs_ndims.into(),
out_dims.base_ptr(ctx, generator).into(),
],
"",
)
.unwrap();
TypedArrayLikeAdapter::from(
out_dims,
Box::new(|_, v| v.into_int_value()),
Box::new(|_, v| v.into()),
)
}
/// Generates a call to `__nac3_ndarray_calc_broadcast_idx`. Returns an [`ArrayAllocaValue`]
/// containing the indices used for accessing `array` corresponding to the index of the broadcasted
/// array `broadcast_idx`.
pub fn call_ndarray_calc_broadcast_index<
'ctx,
G: CodeGenerator + ?Sized,
BroadcastIdx: UntypedArrayLikeAccessor<'ctx>,
>(
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
array: NDArrayValue<'ctx>,
broadcast_idx: &BroadcastIdx,
) -> TypedArrayLikeAdapter<'ctx, IntValue<'ctx>> {
let llvm_i32 = ctx.ctx.i32_type();
let llvm_usize = generator.get_size_type(ctx.ctx);
let llvm_pi32 = llvm_i32.ptr_type(AddressSpace::default());
let llvm_pusize = llvm_usize.ptr_type(AddressSpace::default());
let ndarray_calc_broadcast_fn_name = match llvm_usize.get_bit_width() {
32 => "__nac3_ndarray_calc_broadcast_idx",
64 => "__nac3_ndarray_calc_broadcast_idx64",
bw => codegen_unreachable!(ctx, "Unsupported size type bit width: {}", bw),
};
let ndarray_calc_broadcast_fn =
ctx.module.get_function(ndarray_calc_broadcast_fn_name).unwrap_or_else(|| {
let fn_type = llvm_usize.fn_type(
&[llvm_pusize.into(), llvm_usize.into(), llvm_pi32.into(), llvm_pi32.into()],
false,
);
ctx.module.add_function(ndarray_calc_broadcast_fn_name, fn_type, None)
});
let broadcast_size = broadcast_idx.size(ctx, generator);
let out_idx = ctx.builder.build_array_alloca(llvm_i32, broadcast_size, "").unwrap();
let array_dims = array.dim_sizes().base_ptr(ctx, generator);
let array_ndims = array.load_ndims(ctx);
let broadcast_idx_ptr = unsafe {
broadcast_idx.ptr_offset_unchecked(ctx, generator, &llvm_usize.const_zero(), None)
};
ctx.builder
.build_call(
ndarray_calc_broadcast_fn,
&[array_dims.into(), array_ndims.into(), broadcast_idx_ptr.into(), out_idx.into()],
"",
)
.unwrap();
TypedArrayLikeAdapter::from(
ArraySliceValue::from_ptr_val(out_idx, broadcast_size, None),
Box::new(|_, v| v.into_int_value()),
Box::new(|_, v| v.into()),
)
}

View File

@ -0,0 +1,250 @@
use inkwell::{
values::{BasicValueEnum, IntValue, PointerValue},
AddressSpace,
};
use crate::codegen::{
expr::{create_and_call_function, infer_and_call_function},
irrt::get_usize_dependent_function_name,
types::ProxyType,
values::{ndarray::NDArrayValue, ProxyValue},
CodeGenContext, CodeGenerator,
};
pub fn call_nac3_ndarray_util_assert_shape_no_negative<'ctx, G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
ndims: IntValue<'ctx>,
shape: PointerValue<'ctx>,
) {
let llvm_usize = generator.get_size_type(ctx.ctx);
let llvm_pusize = llvm_usize.ptr_type(AddressSpace::default());
let name = get_usize_dependent_function_name(
generator,
ctx,
"__nac3_ndarray_util_assert_shape_no_negative",
);
create_and_call_function(
ctx,
&name,
Some(llvm_usize.into()),
&[(llvm_usize.into(), ndims.into()), (llvm_pusize.into(), shape.into())],
None,
None,
);
}
pub fn call_nac3_ndarray_util_assert_output_shape_same<'ctx, G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
ndarray_ndims: IntValue<'ctx>,
ndarray_shape: PointerValue<'ctx>,
output_ndims: IntValue<'ctx>,
output_shape: IntValue<'ctx>,
) {
let llvm_usize = generator.get_size_type(ctx.ctx);
let llvm_pusize = llvm_usize.ptr_type(AddressSpace::default());
let name = get_usize_dependent_function_name(
generator,
ctx,
"__nac3_ndarray_util_assert_output_shape_same",
);
create_and_call_function(
ctx,
&name,
Some(llvm_usize.into()),
&[
(llvm_usize.into(), ndarray_ndims.into()),
(llvm_pusize.into(), ndarray_shape.into()),
(llvm_usize.into(), output_ndims.into()),
(llvm_pusize.into(), output_shape.into()),
],
None,
None,
);
}
pub fn call_nac3_ndarray_size<'ctx, G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
ndarray: NDArrayValue<'ctx>,
) -> IntValue<'ctx> {
let llvm_usize = generator.get_size_type(ctx.ctx);
let llvm_ndarray = ndarray.get_type().as_base_type();
let name = get_usize_dependent_function_name(generator, ctx, "__nac3_ndarray_size");
create_and_call_function(
ctx,
&name,
Some(llvm_usize.into()),
&[(llvm_ndarray.into(), ndarray.as_base_value().into())],
Some("size"),
None,
)
.map(BasicValueEnum::into_int_value)
.unwrap()
}
pub fn call_nac3_ndarray_nbytes<'ctx, G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
ndarray: NDArrayValue<'ctx>,
) -> IntValue<'ctx> {
let llvm_usize = generator.get_size_type(ctx.ctx);
let llvm_ndarray = ndarray.get_type().as_base_type();
let name = get_usize_dependent_function_name(generator, ctx, "__nac3_ndarray_nbytes");
create_and_call_function(
ctx,
&name,
Some(llvm_usize.into()),
&[(llvm_ndarray.into(), ndarray.as_base_value().into())],
Some("nbytes"),
None,
)
.map(BasicValueEnum::into_int_value)
.unwrap()
}
pub fn call_nac3_ndarray_len<'ctx, G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
ndarray: NDArrayValue<'ctx>,
) -> IntValue<'ctx> {
let llvm_usize = generator.get_size_type(ctx.ctx);
let llvm_ndarray = ndarray.get_type().as_base_type();
let name = get_usize_dependent_function_name(generator, ctx, "__nac3_ndarray_len");
create_and_call_function(
ctx,
&name,
Some(llvm_usize.into()),
&[(llvm_ndarray.into(), ndarray.as_base_value().into())],
Some("len"),
None,
)
.map(BasicValueEnum::into_int_value)
.unwrap()
}
pub fn call_nac3_ndarray_is_c_contiguous<'ctx, G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
ndarray: NDArrayValue<'ctx>,
) -> IntValue<'ctx> {
let llvm_i1 = ctx.ctx.bool_type();
let llvm_ndarray = ndarray.get_type().as_base_type();
let name = get_usize_dependent_function_name(generator, ctx, "__nac3_ndarray_is_c_contiguous");
create_and_call_function(
ctx,
&name,
Some(llvm_i1.into()),
&[(llvm_ndarray.into(), ndarray.as_base_value().into())],
Some("is_c_contiguous"),
None,
)
.map(BasicValueEnum::into_int_value)
.unwrap()
}
pub fn call_nac3_ndarray_get_nth_pelement<'ctx, G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
ndarray: NDArrayValue<'ctx>,
index: IntValue<'ctx>,
) -> PointerValue<'ctx> {
let llvm_i8 = ctx.ctx.i8_type();
let llvm_pi8 = llvm_i8.ptr_type(AddressSpace::default());
let llvm_usize = generator.get_size_type(ctx.ctx);
let llvm_ndarray = ndarray.get_type().as_base_type();
let name = get_usize_dependent_function_name(generator, ctx, "__nac3_ndarray_get_nth_pelement");
create_and_call_function(
ctx,
&name,
Some(llvm_pi8.into()),
&[(llvm_ndarray.into(), ndarray.as_base_value().into()), (llvm_usize.into(), index.into())],
Some("pelement"),
None,
)
.map(BasicValueEnum::into_pointer_value)
.unwrap()
}
pub fn call_nac3_ndarray_get_pelement_by_indices<'ctx, G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
ndarray: NDArrayValue<'ctx>,
indices: PointerValue<'ctx>,
) -> PointerValue<'ctx> {
let llvm_i8 = ctx.ctx.i8_type();
let llvm_pi8 = llvm_i8.ptr_type(AddressSpace::default());
let llvm_usize = generator.get_size_type(ctx.ctx);
let llvm_pusize = llvm_usize.ptr_type(AddressSpace::default());
let llvm_ndarray = ndarray.get_type().as_base_type();
let name =
get_usize_dependent_function_name(generator, ctx, "__nac3_ndarray_get_pelement_by_indices");
create_and_call_function(
ctx,
&name,
Some(llvm_pi8.into()),
&[
(llvm_ndarray.into(), ndarray.as_base_value().into()),
(llvm_pusize.into(), indices.into()),
],
Some("pelement"),
None,
)
.map(BasicValueEnum::into_pointer_value)
.unwrap()
}
pub fn call_nac3_ndarray_set_strides_by_shape<'ctx, G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
ndarray: NDArrayValue<'ctx>,
) {
let llvm_ndarray = ndarray.get_type().as_base_type();
let name =
get_usize_dependent_function_name(generator, ctx, "__nac3_ndarray_set_strides_by_shape");
create_and_call_function(
ctx,
&name,
None,
&[(llvm_ndarray.into(), ndarray.as_base_value().into())],
None,
None,
);
}
pub fn call_nac3_ndarray_copy_data<'ctx, G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
src_ndarray: NDArrayValue<'ctx>,
dst_ndarray: NDArrayValue<'ctx>,
) {
let name = get_usize_dependent_function_name(generator, ctx, "__nac3_ndarray_copy_data");
infer_and_call_function(
ctx,
&name,
None,
&[src_ndarray.as_base_value().into(), dst_ndarray.as_base_value().into()],
None,
None,
);
}

View File

@ -0,0 +1,29 @@
use crate::codegen::{
expr::infer_and_call_function,
irrt::get_usize_dependent_function_name,
values::{ndarray::NDArrayValue, ArrayLikeValue, ArraySliceValue, ProxyValue},
CodeGenContext, CodeGenerator,
};
pub fn call_nac3_ndarray_index<'ctx, G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
indices: ArraySliceValue<'ctx>,
src_ndarray: NDArrayValue<'ctx>,
dst_ndarray: NDArrayValue<'ctx>,
) {
let name = get_usize_dependent_function_name(generator, ctx, "__nac3_ndarray_index");
infer_and_call_function(
ctx,
&name,
None,
&[
indices.size(ctx, generator).into(),
indices.base_ptr(ctx, generator).into(),
src_ndarray.as_base_value().into(),
dst_ndarray.as_base_value().into(),
],
None,
None,
);
}

View File

@ -0,0 +1,70 @@
use inkwell::{
values::{BasicValueEnum, IntValue},
AddressSpace,
};
use crate::codegen::{
expr::{create_and_call_function, infer_and_call_function},
irrt::get_usize_dependent_function_name,
types::ProxyType,
values::{
ndarray::{NDArrayValue, NDIterValue},
ArrayLikeValue, ArraySliceValue, ProxyValue,
},
CodeGenContext, CodeGenerator,
};
pub fn call_nac3_nditer_initialize<'ctx, G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
iter: NDIterValue<'ctx>,
ndarray: NDArrayValue<'ctx>,
indices: ArraySliceValue<'ctx>,
) {
let llvm_usize = generator.get_size_type(ctx.ctx);
let llvm_pusize = llvm_usize.ptr_type(AddressSpace::default());
let name = get_usize_dependent_function_name(generator, ctx, "__nac3_nditer_initialize");
create_and_call_function(
ctx,
&name,
None,
&[
(iter.get_type().as_base_type().into(), iter.as_base_value().into()),
(ndarray.get_type().as_base_type().into(), ndarray.as_base_value().into()),
(llvm_pusize.into(), indices.base_ptr(ctx, generator).into()),
],
None,
None,
);
}
pub fn call_nac3_nditer_has_element<'ctx, G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
iter: NDIterValue<'ctx>,
) -> IntValue<'ctx> {
let name = get_usize_dependent_function_name(generator, ctx, "__nac3_nditer_has_element");
infer_and_call_function(
ctx,
&name,
Some(ctx.ctx.bool_type().into()),
&[iter.as_base_value().into()],
None,
None,
)
.map(BasicValueEnum::into_int_value)
.unwrap()
}
pub fn call_nac3_nditer_next<'ctx, G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
iter: NDIterValue<'ctx>,
) {
let name = get_usize_dependent_function_name(generator, ctx, "__nac3_nditer_next");
infer_and_call_function(ctx, &name, None, &[iter.as_base_value().into()], None, None);
}

View File

@ -0,0 +1,391 @@
use inkwell::{
types::IntType,
values::{BasicValueEnum, CallSiteValue, IntValue},
AddressSpace, IntPredicate,
};
use itertools::Either;
use crate::codegen::{
llvm_intrinsics,
macros::codegen_unreachable,
stmt::gen_for_callback_incrementing,
values::{
ndarray::NDArrayValue, ArrayLikeIndexer, ArrayLikeValue, ArraySliceValue,
TypedArrayLikeAccessor, TypedArrayLikeAdapter, UntypedArrayLikeAccessor,
},
CodeGenContext, CodeGenerator,
};
pub use basic::*;
pub use indexing::*;
pub use iter::*;
mod basic;
mod indexing;
mod iter;
/// Generates a call to `__nac3_ndarray_calc_size`. Returns an [`IntValue`] representing the
/// calculated total size.
///
/// * `dims` - An [`ArrayLikeIndexer`] containing the size of each dimension.
/// * `range` - The dimension index to begin and end (exclusively) calculating the dimensions for,
/// or [`None`] if starting from the first dimension and ending at the last dimension
/// respectively.
pub fn call_ndarray_calc_size<'ctx, G, Dims>(
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
dims: &Dims,
(begin, end): (Option<IntValue<'ctx>>, Option<IntValue<'ctx>>),
) -> IntValue<'ctx>
where
G: CodeGenerator + ?Sized,
Dims: ArrayLikeIndexer<'ctx>,
{
let llvm_usize = generator.get_size_type(ctx.ctx);
let llvm_pusize = llvm_usize.ptr_type(AddressSpace::default());
let ndarray_calc_size_fn_name = match llvm_usize.get_bit_width() {
32 => "__nac3_ndarray_calc_size",
64 => "__nac3_ndarray_calc_size64",
bw => codegen_unreachable!(ctx, "Unsupported size type bit width: {}", bw),
};
let ndarray_calc_size_fn_t = llvm_usize.fn_type(
&[llvm_pusize.into(), llvm_usize.into(), llvm_usize.into(), llvm_usize.into()],
false,
);
let ndarray_calc_size_fn =
ctx.module.get_function(ndarray_calc_size_fn_name).unwrap_or_else(|| {
ctx.module.add_function(ndarray_calc_size_fn_name, ndarray_calc_size_fn_t, None)
});
let begin = begin.unwrap_or_else(|| llvm_usize.const_zero());
let end = end.unwrap_or_else(|| dims.size(ctx, generator));
ctx.builder
.build_call(
ndarray_calc_size_fn,
&[
dims.base_ptr(ctx, generator).into(),
dims.size(ctx, generator).into(),
begin.into(),
end.into(),
],
"",
)
.map(CallSiteValue::try_as_basic_value)
.map(|v| v.map_left(BasicValueEnum::into_int_value))
.map(Either::unwrap_left)
.unwrap()
}
/// Generates a call to `__nac3_ndarray_calc_nd_indices`. Returns a [`TypeArrayLikeAdpater`]
/// containing `i32` indices of the flattened index.
///
/// * `index` - The index to compute the multidimensional index for.
/// * `ndarray` - LLVM pointer to the `NDArray`. This value must be the LLVM representation of an
/// `NDArray`.
pub fn call_ndarray_calc_nd_indices<'ctx, G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
index: IntValue<'ctx>,
ndarray: NDArrayValue<'ctx>,
) -> TypedArrayLikeAdapter<'ctx, IntValue<'ctx>> {
let llvm_void = ctx.ctx.void_type();
let llvm_i32 = ctx.ctx.i32_type();
let llvm_usize = generator.get_size_type(ctx.ctx);
let llvm_pi32 = llvm_i32.ptr_type(AddressSpace::default());
let llvm_pusize = llvm_usize.ptr_type(AddressSpace::default());
let ndarray_calc_nd_indices_fn_name = match llvm_usize.get_bit_width() {
32 => "__nac3_ndarray_calc_nd_indices",
64 => "__nac3_ndarray_calc_nd_indices64",
bw => codegen_unreachable!(ctx, "Unsupported size type bit width: {}", bw),
};
let ndarray_calc_nd_indices_fn =
ctx.module.get_function(ndarray_calc_nd_indices_fn_name).unwrap_or_else(|| {
let fn_type = llvm_void.fn_type(
&[llvm_usize.into(), llvm_pusize.into(), llvm_usize.into(), llvm_pi32.into()],
false,
);
ctx.module.add_function(ndarray_calc_nd_indices_fn_name, fn_type, None)
});
let ndarray_num_dims = ndarray.load_ndims(ctx);
let ndarray_dims = ndarray.shape();
let indices = ctx.builder.build_array_alloca(llvm_i32, ndarray_num_dims, "").unwrap();
ctx.builder
.build_call(
ndarray_calc_nd_indices_fn,
&[
index.into(),
ndarray_dims.base_ptr(ctx, generator).into(),
ndarray_num_dims.into(),
indices.into(),
],
"",
)
.unwrap();
TypedArrayLikeAdapter::from(
ArraySliceValue::from_ptr_val(indices, ndarray_num_dims, None),
Box::new(|_, v| v.into_int_value()),
Box::new(|_, v| v.into()),
)
}
fn call_ndarray_flatten_index_impl<'ctx, G, Indices>(
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
ndarray: NDArrayValue<'ctx>,
indices: &Indices,
) -> IntValue<'ctx>
where
G: CodeGenerator + ?Sized,
Indices: ArrayLikeIndexer<'ctx>,
{
let llvm_i32 = ctx.ctx.i32_type();
let llvm_usize = generator.get_size_type(ctx.ctx);
let llvm_pi32 = llvm_i32.ptr_type(AddressSpace::default());
let llvm_pusize = llvm_usize.ptr_type(AddressSpace::default());
debug_assert_eq!(
IntType::try_from(indices.element_type(ctx, generator))
.map(IntType::get_bit_width)
.unwrap_or_default(),
llvm_i32.get_bit_width(),
"Expected i32 value for argument `indices` to `call_ndarray_flatten_index_impl`"
);
debug_assert_eq!(
indices.size(ctx, generator).get_type().get_bit_width(),
llvm_usize.get_bit_width(),
"Expected usize integer value for argument `indices_size` to `call_ndarray_flatten_index_impl`"
);
let ndarray_flatten_index_fn_name = match llvm_usize.get_bit_width() {
32 => "__nac3_ndarray_flatten_index",
64 => "__nac3_ndarray_flatten_index64",
bw => codegen_unreachable!(ctx, "Unsupported size type bit width: {}", bw),
};
let ndarray_flatten_index_fn =
ctx.module.get_function(ndarray_flatten_index_fn_name).unwrap_or_else(|| {
let fn_type = llvm_usize.fn_type(
&[llvm_pusize.into(), llvm_usize.into(), llvm_pi32.into(), llvm_usize.into()],
false,
);
ctx.module.add_function(ndarray_flatten_index_fn_name, fn_type, None)
});
let ndarray_num_dims = ndarray.load_ndims(ctx);
let ndarray_dims = ndarray.shape();
let index = ctx
.builder
.build_call(
ndarray_flatten_index_fn,
&[
ndarray_dims.base_ptr(ctx, generator).into(),
ndarray_num_dims.into(),
indices.base_ptr(ctx, generator).into(),
indices.size(ctx, generator).into(),
],
"",
)
.map(CallSiteValue::try_as_basic_value)
.map(|v| v.map_left(BasicValueEnum::into_int_value))
.map(Either::unwrap_left)
.unwrap();
index
}
/// Generates a call to `__nac3_ndarray_flatten_index`. Returns the flattened index for the
/// multidimensional index.
///
/// * `ndarray` - LLVM pointer to the `NDArray`. This value must be the LLVM representation of an
/// `NDArray`.
/// * `indices` - The multidimensional index to compute the flattened index for.
pub fn call_ndarray_flatten_index<'ctx, G, Index>(
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
ndarray: NDArrayValue<'ctx>,
indices: &Index,
) -> IntValue<'ctx>
where
G: CodeGenerator + ?Sized,
Index: ArrayLikeIndexer<'ctx>,
{
call_ndarray_flatten_index_impl(generator, ctx, ndarray, indices)
}
/// Generates a call to `__nac3_ndarray_calc_broadcast`. Returns a tuple containing the number of
/// dimension and size of each dimension of the resultant `ndarray`.
pub fn call_ndarray_calc_broadcast<'ctx, G: CodeGenerator + ?Sized>(
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
lhs: NDArrayValue<'ctx>,
rhs: NDArrayValue<'ctx>,
) -> TypedArrayLikeAdapter<'ctx, IntValue<'ctx>> {
let llvm_usize = generator.get_size_type(ctx.ctx);
let llvm_pusize = llvm_usize.ptr_type(AddressSpace::default());
let ndarray_calc_broadcast_fn_name = match llvm_usize.get_bit_width() {
32 => "__nac3_ndarray_calc_broadcast",
64 => "__nac3_ndarray_calc_broadcast64",
bw => codegen_unreachable!(ctx, "Unsupported size type bit width: {}", bw),
};
let ndarray_calc_broadcast_fn =
ctx.module.get_function(ndarray_calc_broadcast_fn_name).unwrap_or_else(|| {
let fn_type = llvm_usize.fn_type(
&[
llvm_pusize.into(),
llvm_usize.into(),
llvm_pusize.into(),
llvm_usize.into(),
llvm_pusize.into(),
],
false,
);
ctx.module.add_function(ndarray_calc_broadcast_fn_name, fn_type, None)
});
let lhs_ndims = lhs.load_ndims(ctx);
let rhs_ndims = rhs.load_ndims(ctx);
let min_ndims = llvm_intrinsics::call_int_umin(ctx, lhs_ndims, rhs_ndims, None);
gen_for_callback_incrementing(
generator,
ctx,
None,
llvm_usize.const_zero(),
(min_ndims, false),
|generator, ctx, _, idx| {
let idx = ctx.builder.build_int_sub(min_ndims, idx, "").unwrap();
let (lhs_dim_sz, rhs_dim_sz) = unsafe {
(
lhs.shape().get_typed_unchecked(ctx, generator, &idx, None),
rhs.shape().get_typed_unchecked(ctx, generator, &idx, None),
)
};
let llvm_usize_const_one = llvm_usize.const_int(1, false);
let lhs_eqz = ctx
.builder
.build_int_compare(IntPredicate::EQ, lhs_dim_sz, llvm_usize_const_one, "")
.unwrap();
let rhs_eqz = ctx
.builder
.build_int_compare(IntPredicate::EQ, rhs_dim_sz, llvm_usize_const_one, "")
.unwrap();
let lhs_or_rhs_eqz = ctx.builder.build_or(lhs_eqz, rhs_eqz, "").unwrap();
let lhs_eq_rhs = ctx
.builder
.build_int_compare(IntPredicate::EQ, lhs_dim_sz, rhs_dim_sz, "")
.unwrap();
let is_compatible = ctx.builder.build_or(lhs_or_rhs_eqz, lhs_eq_rhs, "").unwrap();
ctx.make_assert(
generator,
is_compatible,
"0:ValueError",
"operands could not be broadcast together",
[None, None, None],
ctx.current_loc,
);
Ok(())
},
llvm_usize.const_int(1, false),
)
.unwrap();
let max_ndims = llvm_intrinsics::call_int_umax(ctx, lhs_ndims, rhs_ndims, None);
let lhs_dims = lhs.shape().base_ptr(ctx, generator);
let lhs_ndims = lhs.load_ndims(ctx);
let rhs_dims = rhs.shape().base_ptr(ctx, generator);
let rhs_ndims = rhs.load_ndims(ctx);
let out_dims = ctx.builder.build_array_alloca(llvm_usize, max_ndims, "").unwrap();
let out_dims = ArraySliceValue::from_ptr_val(out_dims, max_ndims, None);
ctx.builder
.build_call(
ndarray_calc_broadcast_fn,
&[
lhs_dims.into(),
lhs_ndims.into(),
rhs_dims.into(),
rhs_ndims.into(),
out_dims.base_ptr(ctx, generator).into(),
],
"",
)
.unwrap();
TypedArrayLikeAdapter::from(
out_dims,
Box::new(|_, v| v.into_int_value()),
Box::new(|_, v| v.into()),
)
}
/// Generates a call to `__nac3_ndarray_calc_broadcast_idx`. Returns an [`ArrayAllocaValue`]
/// containing the indices used for accessing `array` corresponding to the index of the broadcasted
/// array `broadcast_idx`.
pub fn call_ndarray_calc_broadcast_index<
'ctx,
G: CodeGenerator + ?Sized,
BroadcastIdx: UntypedArrayLikeAccessor<'ctx>,
>(
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
array: NDArrayValue<'ctx>,
broadcast_idx: &BroadcastIdx,
) -> TypedArrayLikeAdapter<'ctx, IntValue<'ctx>> {
let llvm_i32 = ctx.ctx.i32_type();
let llvm_usize = generator.get_size_type(ctx.ctx);
let llvm_pi32 = llvm_i32.ptr_type(AddressSpace::default());
let llvm_pusize = llvm_usize.ptr_type(AddressSpace::default());
let ndarray_calc_broadcast_fn_name = match llvm_usize.get_bit_width() {
32 => "__nac3_ndarray_calc_broadcast_idx",
64 => "__nac3_ndarray_calc_broadcast_idx64",
bw => codegen_unreachable!(ctx, "Unsupported size type bit width: {}", bw),
};
let ndarray_calc_broadcast_fn =
ctx.module.get_function(ndarray_calc_broadcast_fn_name).unwrap_or_else(|| {
let fn_type = llvm_usize.fn_type(
&[llvm_pusize.into(), llvm_usize.into(), llvm_pi32.into(), llvm_pi32.into()],
false,
);
ctx.module.add_function(ndarray_calc_broadcast_fn_name, fn_type, None)
});
let broadcast_size = broadcast_idx.size(ctx, generator);
let out_idx = ctx.builder.build_array_alloca(llvm_i32, broadcast_size, "").unwrap();
let array_dims = array.shape().base_ptr(ctx, generator);
let array_ndims = array.load_ndims(ctx);
let broadcast_idx_ptr = unsafe {
broadcast_idx.ptr_offset_unchecked(ctx, generator, &llvm_usize.const_zero(), None)
};
ctx.builder
.build_call(
ndarray_calc_broadcast_fn,
&[array_dims.into(), array_ndims.into(), broadcast_idx_ptr.into(), out_idx.into()],
"",
)
.unwrap();
TypedArrayLikeAdapter::from(
ArraySliceValue::from_ptr_val(out_idx, broadcast_size, None),
Box::new(|_, v| v.into_int_value()),
Box::new(|_, v| v.into()),
)
}

View File

@ -0,0 +1,42 @@
use inkwell::{
values::{BasicValueEnum, CallSiteValue, IntValue},
IntPredicate,
};
use itertools::Either;
use crate::codegen::{CodeGenContext, CodeGenerator};
pub fn calculate_len_for_slice_range<'ctx, G: CodeGenerator + ?Sized>(
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
start: IntValue<'ctx>,
end: IntValue<'ctx>,
step: IntValue<'ctx>,
) -> IntValue<'ctx> {
const SYMBOL: &str = "__nac3_range_slice_len";
let len_func = ctx.module.get_function(SYMBOL).unwrap_or_else(|| {
let i32_t = ctx.ctx.i32_type();
let fn_t = i32_t.fn_type(&[i32_t.into(), i32_t.into(), i32_t.into()], false);
ctx.module.add_function(SYMBOL, fn_t, None)
});
// assert step != 0, throw exception if not
let not_zero = ctx
.builder
.build_int_compare(IntPredicate::NE, step, step.get_type().const_zero(), "range_step_ne")
.unwrap();
ctx.make_assert(
generator,
not_zero,
"0:ValueError",
"step must not be zero",
[None, None, None],
ctx.current_loc,
);
ctx.builder
.build_call(len_func, &[start.into(), end.into(), step.into()], "calc_len")
.map(CallSiteValue::try_as_basic_value)
.map(|v| v.map_left(BasicValueEnum::into_int_value))
.map(Either::unwrap_left)
.unwrap()
}

View File

@ -0,0 +1,39 @@
use inkwell::values::{BasicValueEnum, CallSiteValue, IntValue};
use itertools::Either;
use nac3parser::ast::Expr;
use crate::{
codegen::{CodeGenContext, CodeGenerator},
typecheck::typedef::Type,
};
/// this function allows index out of range, since python
/// allows index out of range in slice (`a = [1,2,3]; a[1:10] == [2,3]`).
pub fn handle_slice_index_bound<'ctx, G: CodeGenerator>(
i: &Expr<Option<Type>>,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
length: IntValue<'ctx>,
) -> Result<Option<IntValue<'ctx>>, String> {
const SYMBOL: &str = "__nac3_slice_index_bound";
let func = ctx.module.get_function(SYMBOL).unwrap_or_else(|| {
let i32_t = ctx.ctx.i32_type();
let fn_t = i32_t.fn_type(&[i32_t.into(), i32_t.into()], false);
ctx.module.add_function(SYMBOL, fn_t, None)
});
let i = if let Some(v) = generator.gen_expr(ctx, i)? {
v.to_basic_value_enum(ctx, generator, i.custom.unwrap())?
} else {
return Ok(None);
};
Ok(Some(
ctx.builder
.build_call(func, &[i.into(), length.into()], "bounded_ind")
.map(CallSiteValue::try_as_basic_value)
.map(|v| v.map_left(BasicValueEnum::into_int_value))
.map(Either::unwrap_left)
.unwrap(),
))
}

View File

@ -1,12 +1,14 @@
use crate::codegen::CodeGenContext;
use inkwell::context::Context;
use inkwell::intrinsics::Intrinsic;
use inkwell::types::AnyTypeEnum::IntType;
use inkwell::types::FloatType;
use inkwell::values::{BasicValueEnum, CallSiteValue, FloatValue, IntValue, PointerValue};
use inkwell::AddressSpace;
use inkwell::{
context::Context,
intrinsics::Intrinsic,
types::{AnyTypeEnum::IntType, FloatType},
values::{BasicValueEnum, CallSiteValue, FloatValue, IntValue, PointerValue},
AddressSpace,
};
use itertools::Either;
use super::CodeGenContext;
/// Returns the string representation for the floating-point type `ft` when used in intrinsic
/// functions.
fn get_float_intrinsic_repr(ctx: &Context, ft: FloatType) -> &'static str {
@ -183,7 +185,7 @@ pub fn call_memcpy_generic<'ctx>(
dest
} else {
ctx.builder
.build_bitcast(dest, llvm_p0i8, "")
.build_bit_cast(dest, llvm_p0i8, "")
.map(BasicValueEnum::into_pointer_value)
.unwrap()
};
@ -191,7 +193,7 @@ pub fn call_memcpy_generic<'ctx>(
src
} else {
ctx.builder
.build_bitcast(src, llvm_p0i8, "")
.build_bit_cast(src, llvm_p0i8, "")
.map(BasicValueEnum::into_pointer_value)
.unwrap()
};
@ -199,6 +201,49 @@ pub fn call_memcpy_generic<'ctx>(
call_memcpy(ctx, dest, src, len, is_volatile);
}
/// Invokes the `llvm.memcpy` intrinsic.
///
/// Unlike [`call_memcpy`], this function accepts any type of pointer value. If `dest` or `src` is
/// not a pointer to an integer, the pointer(s) will be cast to `i8*` before invoking `memcpy`.
/// Moreover, `len` now refers to the number of elements to copy (rather than number of bytes to
/// copy).
pub fn call_memcpy_generic_array<'ctx>(
ctx: &CodeGenContext<'ctx, '_>,
dest: PointerValue<'ctx>,
src: PointerValue<'ctx>,
len: IntValue<'ctx>,
is_volatile: IntValue<'ctx>,
) {
let llvm_i8 = ctx.ctx.i8_type();
let llvm_p0i8 = llvm_i8.ptr_type(AddressSpace::default());
let llvm_sizeof_expr_t = llvm_i8.size_of().get_type();
let dest_elem_t = dest.get_type().get_element_type();
let src_elem_t = src.get_type().get_element_type();
let dest = if matches!(dest_elem_t, IntType(t) if t.get_bit_width() == 8) {
dest
} else {
ctx.builder
.build_bit_cast(dest, llvm_p0i8, "")
.map(BasicValueEnum::into_pointer_value)
.unwrap()
};
let src = if matches!(src_elem_t, IntType(t) if t.get_bit_width() == 8) {
src
} else {
ctx.builder
.build_bit_cast(src, llvm_p0i8, "")
.map(BasicValueEnum::into_pointer_value)
.unwrap()
};
let len = ctx.builder.build_int_z_extend_or_bit_cast(len, llvm_sizeof_expr_t, "").unwrap();
let len = ctx.builder.build_int_mul(len, src_elem_t.size_of().unwrap(), "").unwrap();
call_memcpy(ctx, dest, src, len, is_volatile);
}
/// Macro to find and generate build call for llvm intrinsic (body of llvm intrinsic function)
///
/// Arguments:
@ -341,3 +386,25 @@ pub fn call_float_powi<'ctx>(
.map(Either::unwrap_left)
.unwrap()
}
/// Invokes the [`llvm.ctpop`](https://llvm.org/docs/LangRef.html#llvm-ctpop-intrinsic) intrinsic.
pub fn call_int_ctpop<'ctx>(
ctx: &CodeGenContext<'ctx, '_>,
src: IntValue<'ctx>,
name: Option<&str>,
) -> IntValue<'ctx> {
const FN_NAME: &str = "llvm.ctpop";
let llvm_src_t = src.get_type();
let intrinsic_fn = Intrinsic::find(FN_NAME)
.and_then(|intrinsic| intrinsic.get_declaration(&ctx.module, &[llvm_src_t.into()]))
.unwrap();
ctx.builder
.build_call(intrinsic_fn, &[src.into()], name.unwrap_or_default())
.map(CallSiteValue::try_as_basic_value)
.map(|v| v.map_left(BasicValueEnum::into_int_value))
.map(Either::unwrap_left)
.unwrap()
}

View File

@ -1,12 +1,12 @@
use crate::{
codegen::classes::{ListType, NDArrayType, ProxyType, RangeType},
symbol_resolver::{StaticValue, SymbolResolver},
toplevel::{helper::PrimDef, numpy::unpack_ndarray_var_tys, TopLevelContext, TopLevelDef},
typecheck::{
type_inferencer::{CodeLocation, PrimitiveStore},
typedef::{CallId, FuncArg, Type, TypeEnum, Unifier},
use std::{
collections::{HashMap, HashSet},
sync::{
atomic::{AtomicBool, Ordering},
Arc,
},
thread,
};
use crossbeam::channel::{unbounded, Receiver, Sender};
use inkwell::{
attributes::{Attribute, AttributeLoc},
@ -24,17 +24,27 @@ use inkwell::{
AddressSpace, IntPredicate, OptimizationLevel,
};
use itertools::Itertools;
use nac3parser::ast::{Location, Stmt, StrRef};
use parking_lot::{Condvar, Mutex};
use std::collections::{HashMap, HashSet};
use std::sync::{
atomic::{AtomicBool, Ordering},
Arc,
use nac3parser::ast::{Location, Stmt, StrRef};
use crate::{
symbol_resolver::{StaticValue, SymbolResolver},
toplevel::{
helper::{extract_ndims, PrimDef},
numpy::unpack_ndarray_var_tys,
TopLevelContext, TopLevelDef,
},
typecheck::{
type_inferencer::{CodeLocation, PrimitiveStore},
typedef::{CallId, FuncArg, Type, TypeEnum, Unifier},
},
};
use std::thread;
use concrete_type::{ConcreteType, ConcreteTypeEnum, ConcreteTypeStore};
pub use generator::{CodeGenerator, DefaultCodeGenerator};
use types::{ndarray::NDArrayType, ListType, ProxyType, RangeType};
pub mod builtin_fns;
pub mod classes;
pub mod concrete_type;
pub mod expr;
pub mod extern_fns;
@ -43,13 +53,12 @@ pub mod irrt;
pub mod llvm_intrinsics;
pub mod numpy;
pub mod stmt;
pub mod types;
pub mod values;
#[cfg(test)]
mod test;
use concrete_type::{ConcreteType, ConcreteTypeEnum, ConcreteTypeStore};
pub use generator::{CodeGenerator, DefaultCodeGenerator};
mod macros {
/// Codegen-variant of [`std::unreachable`] which accepts an instance of [`CodeGenContext`] as
/// its first argument to provide Python source information to indicate the codegen location
@ -505,12 +514,13 @@ fn get_llvm_type<'ctx, G: CodeGenerator + ?Sized>(
}
TObj { obj_id, .. } if *obj_id == PrimDef::NDArray.id() => {
let (dtype, _) = unpack_ndarray_var_tys(unifier, ty);
let (dtype, ndims) = unpack_ndarray_var_tys(unifier, ty);
let ndims = extract_ndims(unifier, ndims);
let element_type = get_llvm_type(
ctx, module, generator, unifier, top_level, type_cache, dtype,
);
NDArrayType::new(generator, ctx, element_type).as_base_type().into()
NDArrayType::new(generator, ctx, element_type, Some(ndims)).as_base_type().into()
}
_ => unreachable!(
@ -848,10 +858,9 @@ pub fn gen_func_impl<
builder.position_at_end(init_bb);
let body_bb = context.append_basic_block(fn_val, "body");
// Store non-vararg argument values into local variables
let mut var_assignment = HashMap::new();
let offset = u32::from(has_sret);
// Store non-vararg argument values into local variables
for (n, arg) in args.iter().enumerate().filter(|(_, arg)| !arg.is_vararg) {
let param = fn_val.get_nth_param((n as u32) + offset).unwrap();
let local_type = get_llvm_type(
@ -1115,3 +1124,106 @@ fn gen_in_range_check<'ctx>(
fn get_va_count_arg_name(arg_name: StrRef) -> StrRef {
format!("__{}_va_count", &arg_name).into()
}
/// Returns the alignment of the type.
///
/// This is necessary as `get_alignment` is not implemented as part of [`BasicType`].
pub fn get_type_alignment<'ctx>(ty: impl Into<BasicTypeEnum<'ctx>>) -> IntValue<'ctx> {
match ty.into() {
BasicTypeEnum::ArrayType(ty) => ty.get_alignment(),
BasicTypeEnum::FloatType(ty) => ty.get_alignment(),
BasicTypeEnum::IntType(ty) => ty.get_alignment(),
BasicTypeEnum::PointerType(ty) => ty.get_alignment(),
BasicTypeEnum::StructType(ty) => ty.get_alignment(),
BasicTypeEnum::VectorType(ty) => ty.get_alignment(),
}
}
/// Inserts an `alloca` instruction with allocation `size` given in bytes and the alignment of the
/// given type.
///
/// The returned [`PointerValue`] will have a type of `i8*`, a size of at least `size`, and will be
/// aligned with the alignment of `align_ty`.
pub fn type_aligned_alloca<'ctx, G: CodeGenerator + ?Sized>(
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
align_ty: impl Into<BasicTypeEnum<'ctx>>,
size: IntValue<'ctx>,
name: Option<&str>,
) -> PointerValue<'ctx> {
/// Round `val` up to its modulo `power_of_two`.
fn round_up<'ctx>(
ctx: &CodeGenContext<'ctx, '_>,
val: IntValue<'ctx>,
power_of_two: IntValue<'ctx>,
) -> IntValue<'ctx> {
debug_assert_eq!(
val.get_type().get_bit_width(),
power_of_two.get_type().get_bit_width(),
"`val` ({}) and `power_of_two` ({}) must be the same type",
val.get_type(),
power_of_two.get_type(),
);
let llvm_val_t = val.get_type();
let max_rem =
ctx.builder.build_int_sub(power_of_two, llvm_val_t.const_int(1, false), "").unwrap();
ctx.builder
.build_and(
ctx.builder.build_int_add(val, max_rem, "").unwrap(),
ctx.builder.build_not(max_rem, "").unwrap(),
"",
)
.unwrap()
}
let llvm_i8 = ctx.ctx.i8_type();
let llvm_pi8 = llvm_i8.ptr_type(AddressSpace::default());
let llvm_usize = generator.get_size_type(ctx.ctx);
let align_ty = align_ty.into();
let size = ctx.builder.build_int_truncate_or_bit_cast(size, llvm_usize, "").unwrap();
debug_assert_eq!(
size.get_type().get_bit_width(),
llvm_usize.get_bit_width(),
"Expected size_t ({}) for parameter `size` of `aligned_alloca`, got {}",
llvm_usize,
size.get_type(),
);
let alignment = get_type_alignment(align_ty);
let alignment = ctx.builder.build_int_truncate_or_bit_cast(alignment, llvm_usize, "").unwrap();
if ctx.registry.llvm_options.opt_level == OptimizationLevel::None {
let alignment_bitcount = llvm_intrinsics::call_int_ctpop(ctx, alignment, None);
ctx.make_assert(
generator,
ctx.builder
.build_int_compare(
IntPredicate::EQ,
alignment_bitcount,
alignment_bitcount.get_type().const_int(1, false),
"",
)
.unwrap(),
"0:AssertionError",
"Expected power-of-two alignment for aligned_alloca, got {0}",
[Some(alignment), None, None],
ctx.current_loc,
);
}
let buffer_size = round_up(ctx, size, alignment);
let aligned_slices = ctx.builder.build_int_unsigned_div(buffer_size, alignment, "").unwrap();
// Just to be absolutely sure, alloca in [i8 x alignment] slices
let buffer = ctx.builder.build_array_alloca(align_ty, aligned_slices, "").unwrap();
ctx.builder
.build_bit_cast(buffer, llvm_pi8, name.unwrap_or_default())
.map(BasicValueEnum::into_pointer_value)
.unwrap()
}

View File

@ -1,63 +1,40 @@
use crate::{
codegen::{
classes::{
ArrayLikeIndexer, ArrayLikeValue, ListType, ListValue, NDArrayType, NDArrayValue,
ProxyType, ProxyValue, TypedArrayLikeAccessor, TypedArrayLikeAdapter,
TypedArrayLikeMutator, UntypedArrayLikeAccessor, UntypedArrayLikeMutator,
},
use inkwell::{
types::{AnyTypeEnum, BasicType, BasicTypeEnum, PointerType},
values::{BasicValue, BasicValueEnum, IntValue, PointerValue},
AddressSpace, IntPredicate, OptimizationLevel,
};
use itertools::Itertools;
use nac3parser::ast::{Operator, StrRef};
use super::{
expr::gen_binop_expr_with_values,
irrt::{
calculate_len_for_slice_range, call_ndarray_calc_broadcast,
call_ndarray_calc_broadcast_index, call_ndarray_calc_nd_indices,
call_ndarray_calc_size,
calculate_len_for_slice_range,
ndarray::{
call_ndarray_calc_broadcast, call_ndarray_calc_broadcast_index,
call_ndarray_calc_nd_indices, call_ndarray_calc_size,
},
},
llvm_intrinsics::{self, call_memcpy_generic},
macros::codegen_unreachable,
stmt::{gen_for_callback_incrementing, gen_for_range_callback, gen_if_else_expr_callback},
types::{ndarray::NDArrayType, ListType, ProxyType},
values::{
ndarray::NDArrayValue, ArrayLikeIndexer, ArrayLikeValue, ListValue, ProxyValue,
TypedArrayLikeAccessor, TypedArrayLikeAdapter, TypedArrayLikeMutator,
UntypedArrayLikeAccessor, UntypedArrayLikeMutator,
},
CodeGenContext, CodeGenerator,
},
};
use crate::{
symbol_resolver::ValueEnum,
toplevel::{
helper::PrimDef,
numpy::{make_ndarray_ty, unpack_ndarray_var_tys},
DefinitionId,
},
toplevel::{helper::PrimDef, numpy::unpack_ndarray_var_tys, DefinitionId},
typecheck::{
magic_methods::Binop,
typedef::{FunSignature, Type, TypeEnum},
},
};
use inkwell::{
types::BasicType,
values::{BasicValueEnum, IntValue, PointerValue},
AddressSpace, IntPredicate, OptimizationLevel,
};
use inkwell::{
types::{AnyTypeEnum, BasicTypeEnum, PointerType},
values::BasicValue,
};
use nac3parser::ast::{Operator, StrRef};
/// Creates an uninitialized `NDArray` instance.
fn create_ndarray_uninitialized<'ctx, G: CodeGenerator + ?Sized>(
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
elem_ty: Type,
) -> Result<NDArrayValue<'ctx>, String> {
let ndarray_ty = make_ndarray_ty(&mut ctx.unifier, &ctx.primitives, Some(elem_ty), None);
let llvm_usize = generator.get_size_type(ctx.ctx);
let llvm_ndarray_t = ctx
.get_llvm_type(generator, ndarray_ty)
.into_pointer_type()
.get_element_type()
.into_struct_type();
let ndarray = generator.gen_var_alloc(ctx, llvm_ndarray_t.into(), None)?;
Ok(NDArrayValue::from_ptr_val(ndarray, llvm_usize, None))
}
/// Creates an `NDArray` instance from a dynamic shape.
///
@ -84,6 +61,7 @@ where
) -> Result<IntValue<'ctx>, String>,
{
let llvm_usize = generator.get_size_type(ctx.ctx);
let llvm_elem_ty = ctx.get_llvm_type(generator, elem_ty);
// Assert that all dimensions are non-negative
let shape_len = shape_len_fn(generator, ctx, shape)?;
@ -116,20 +94,17 @@ where
ctx.current_loc,
);
// TODO: Disallow dim_sz > u32_MAX
// TODO: Disallow shape > u32_MAX
Ok(())
},
llvm_usize.const_int(1, false),
)?;
let ndarray = create_ndarray_uninitialized(generator, ctx, elem_ty)?;
let num_dims = shape_len_fn(generator, ctx, shape)?;
ndarray.store_ndims(ctx, generator, num_dims);
let ndarray_num_dims = ndarray.load_ndims(ctx);
ndarray.create_dim_sizes(ctx, llvm_usize, ndarray_num_dims);
let ndarray = NDArrayType::new(generator, ctx.ctx, llvm_elem_ty, None)
.construct_dyn_ndims(generator, ctx, num_dims, None);
// Copy the dimension sizes from shape to ndarray.dims
let shape_len = shape_len_fn(generator, ctx, shape)?;
@ -145,7 +120,7 @@ where
let shape_dim = ctx.builder.build_int_z_extend(shape_dim, llvm_usize, "").unwrap();
let ndarray_pdim =
unsafe { ndarray.dim_sizes().ptr_offset_unchecked(ctx, generator, &i, None) };
unsafe { ndarray.shape().ptr_offset_unchecked(ctx, generator, &i, None) };
ctx.builder.build_store(ndarray_pdim, shape_dim).unwrap();
@ -154,7 +129,7 @@ where
llvm_usize.const_int(1, false),
)?;
let ndarray = ndarray_init_data(generator, ctx, elem_ty, ndarray);
unsafe { ndarray.create_data(generator, ctx) };
Ok(ndarray)
}
@ -187,57 +162,18 @@ pub fn create_ndarray_const_shape<'ctx, G: CodeGenerator + ?Sized>(
ctx.current_loc,
);
// TODO: Disallow dim_sz > u32_MAX
// TODO: Disallow shape > u32_MAX
}
let ndarray = create_ndarray_uninitialized(generator, ctx, elem_ty)?;
let llvm_dtype = ctx.get_llvm_type(generator, elem_ty);
let num_dims = llvm_usize.const_int(shape.len() as u64, false);
ndarray.store_ndims(ctx, generator, num_dims);
let ndarray_num_dims = ndarray.load_ndims(ctx);
ndarray.create_dim_sizes(ctx, llvm_usize, ndarray_num_dims);
for (i, &shape_dim) in shape.iter().enumerate() {
let shape_dim = ctx.builder.build_int_z_extend(shape_dim, llvm_usize, "").unwrap();
let ndarray_dim = unsafe {
ndarray.dim_sizes().ptr_offset_unchecked(
ctx,
generator,
&llvm_usize.const_int(i as u64, true),
None,
)
};
ctx.builder.build_store(ndarray_dim, shape_dim).unwrap();
}
let ndarray = ndarray_init_data(generator, ctx, elem_ty, ndarray);
let ndarray = NDArrayType::new(generator, ctx.ctx, llvm_dtype, Some(shape.len() as u64))
.construct_dyn_shape(generator, ctx, shape, None);
unsafe { ndarray.create_data(generator, ctx) };
Ok(ndarray)
}
/// Initializes the `data` field of [`NDArrayValue`] based on the `ndims` and `dim_sz` fields.
fn ndarray_init_data<'ctx, G: CodeGenerator + ?Sized>(
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
elem_ty: Type,
ndarray: NDArrayValue<'ctx>,
) -> NDArrayValue<'ctx> {
let llvm_ndarray_data_t = ctx.get_llvm_type(generator, elem_ty).as_basic_type_enum();
assert!(llvm_ndarray_data_t.is_sized());
let ndarray_num_elems = call_ndarray_calc_size(
generator,
ctx,
&ndarray.dim_sizes().as_slice_value(ctx, generator),
(None, None),
);
ndarray.create_data(ctx, llvm_ndarray_data_t, ndarray_num_elems);
ndarray
}
fn ndarray_zero_value<'ctx, G: CodeGenerator + ?Sized>(
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
@ -316,11 +252,11 @@ fn call_ndarray_empty_impl<'ctx, G: CodeGenerator + ?Sized>(
match shape {
BasicValueEnum::PointerValue(shape_list_ptr)
if ListValue::is_instance(shape_list_ptr, llvm_usize).is_ok() =>
if ListValue::is_representable(shape_list_ptr, llvm_usize).is_ok() =>
{
// 1. A list of ints; e.g., `np.empty([600, 800, 3])`
let shape_list = ListValue::from_ptr_val(shape_list_ptr, llvm_usize, None);
let shape_list = ListValue::from_pointer_value(shape_list_ptr, llvm_usize, None);
create_ndarray_dyn_shape(
generator,
ctx,
@ -339,20 +275,24 @@ fn call_ndarray_empty_impl<'ctx, G: CodeGenerator + ?Sized>(
// Get the length/size of the tuple, which also happens to be the value of `ndims`.
let ndims = shape_tuple.get_type().count_fields();
let mut shape = Vec::with_capacity(ndims as usize);
for dim_i in 0..ndims {
let dim = ctx
.builder
let shape = (0..ndims)
.map(|dim_i| {
ctx.builder
.build_extract_value(shape_tuple, dim_i, format!("dim{dim_i}").as_str())
.map(BasicValueEnum::into_int_value)
.map(|v| {
ctx.builder.build_int_z_extend_or_bit_cast(v, llvm_usize, "").unwrap()
})
.unwrap()
.into_int_value();
})
.collect_vec();
shape.push(dim);
}
create_ndarray_const_shape(generator, ctx, elem_ty, shape.as_slice())
}
BasicValueEnum::IntValue(shape_int) => {
// 3. A scalar int; e.g., `np.empty(3)`, this is functionally equivalent to `np.empty([3])`
let shape_int =
ctx.builder.build_int_z_extend_or_bit_cast(shape_int, llvm_usize, "").unwrap();
create_ndarray_const_shape(generator, ctx, elem_ty, &[shape_int])
}
@ -381,7 +321,7 @@ where
let ndarray_num_elems = call_ndarray_calc_size(
generator,
ctx,
&ndarray.dim_sizes().as_slice_value(ctx, generator),
&ndarray.shape().as_slice_value(ctx, generator),
(None, None),
);
@ -475,8 +415,8 @@ fn ndarray_broadcast_fill<'ctx, 'a, G, ValueFn>(
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, 'a>,
res: NDArrayValue<'ctx>,
lhs: (BasicValueEnum<'ctx>, bool),
rhs: (BasicValueEnum<'ctx>, bool),
(lhs_ty, lhs_val, lhs_scalar): (Type, BasicValueEnum<'ctx>, bool),
(rhs_ty, rhs_val, rhs_scalar): (Type, BasicValueEnum<'ctx>, bool),
value_fn: ValueFn,
) -> Result<NDArrayValue<'ctx>, String>
where
@ -487,11 +427,6 @@ where
(BasicValueEnum<'ctx>, BasicValueEnum<'ctx>),
) -> Result<BasicValueEnum<'ctx>, String>,
{
let llvm_usize = generator.get_size_type(ctx.ctx);
let (lhs_val, lhs_scalar) = lhs;
let (rhs_val, rhs_scalar) = rhs;
assert!(
!(lhs_scalar && rhs_scalar),
"One of the operands must be a ndarray instance: `{}`, `{}`",
@ -501,12 +436,14 @@ where
// Assert that all ndarray operands are broadcastable to the target size
if !lhs_scalar {
let lhs_val = NDArrayValue::from_ptr_val(lhs_val.into_pointer_value(), llvm_usize, None);
let lhs_val = NDArrayType::from_unifier_type(generator, ctx, lhs_ty)
.map_value(lhs_val.into_pointer_value(), None);
ndarray_assert_is_broadcastable(generator, ctx, res, lhs_val);
}
if !rhs_scalar {
let rhs_val = NDArrayValue::from_ptr_val(rhs_val.into_pointer_value(), llvm_usize, None);
let rhs_val = NDArrayType::from_unifier_type(generator, ctx, rhs_ty)
.map_value(rhs_val.into_pointer_value(), None);
ndarray_assert_is_broadcastable(generator, ctx, res, rhs_val);
}
@ -514,7 +451,8 @@ where
let lhs_elem = if lhs_scalar {
lhs_val
} else {
let lhs = NDArrayValue::from_ptr_val(lhs_val.into_pointer_value(), llvm_usize, None);
let lhs = NDArrayType::from_unifier_type(generator, ctx, lhs_ty)
.map_value(lhs_val.into_pointer_value(), None);
let lhs_idx = call_ndarray_calc_broadcast_index(generator, ctx, lhs, idx);
unsafe { lhs.data().get_unchecked(ctx, generator, &lhs_idx, None) }
@ -523,7 +461,8 @@ where
let rhs_elem = if rhs_scalar {
rhs_val
} else {
let rhs = NDArrayValue::from_ptr_val(rhs_val.into_pointer_value(), llvm_usize, None);
let rhs = NDArrayType::from_unifier_type(generator, ctx, rhs_ty)
.map_value(rhs_val.into_pointer_value(), None);
let rhs_idx = call_ndarray_calc_broadcast_index(generator, ctx, rhs, idx);
unsafe { rhs.data().get_unchecked(ctx, generator, &rhs_idx, None) }
@ -649,11 +588,15 @@ fn llvm_ndlist_get_ndims<'ctx, G: CodeGenerator + ?Sized>(
let ndims = llvm_usize.const_int(1, false);
match list_elem_ty {
AnyTypeEnum::PointerType(ptr_ty) if ListType::is_type(ptr_ty, llvm_usize).is_ok() => {
AnyTypeEnum::PointerType(ptr_ty)
if ListType::is_representable(ptr_ty, llvm_usize).is_ok() =>
{
ndims.const_add(llvm_ndlist_get_ndims(generator, ctx, ptr_ty))
}
AnyTypeEnum::PointerType(ptr_ty) if NDArrayType::is_type(ptr_ty, llvm_usize).is_ok() => {
AnyTypeEnum::PointerType(ptr_ty)
if NDArrayType::is_representable(ptr_ty, llvm_usize).is_ok() =>
{
todo!("Getting ndims for list[ndarray] not supported")
}
@ -665,16 +608,18 @@ fn llvm_ndlist_get_ndims<'ctx, G: CodeGenerator + ?Sized>(
fn llvm_arraylike_get_ndims<'ctx, G: CodeGenerator + ?Sized>(
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
value: BasicValueEnum<'ctx>,
(ty, value): (Type, BasicValueEnum<'ctx>),
) -> IntValue<'ctx> {
let llvm_usize = generator.get_size_type(ctx.ctx);
match value {
BasicValueEnum::PointerValue(v) if NDArrayValue::is_instance(v, llvm_usize).is_ok() => {
NDArrayValue::from_ptr_val(v, llvm_usize, None).load_ndims(ctx)
BasicValueEnum::PointerValue(v)
if NDArrayValue::is_representable(v, llvm_usize).is_ok() =>
{
NDArrayType::from_unifier_type(generator, ctx, ty).map_value(v, None).load_ndims(ctx)
}
BasicValueEnum::PointerValue(v) if ListValue::is_instance(v, llvm_usize).is_ok() => {
BasicValueEnum::PointerValue(v) if ListValue::is_representable(v, llvm_usize).is_ok() => {
llvm_ndlist_get_ndims(generator, ctx, v.get_type())
}
@ -686,7 +631,6 @@ fn llvm_arraylike_get_ndims<'ctx, G: CodeGenerator + ?Sized>(
fn ndarray_from_ndlist_impl<'ctx, G: CodeGenerator + ?Sized>(
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
elem_ty: Type,
(dst_arr, dst_slice_ptr): (NDArrayValue<'ctx>, PointerValue<'ctx>),
src_lst: ListValue<'ctx>,
dim: u64,
@ -697,13 +641,15 @@ fn ndarray_from_ndlist_impl<'ctx, G: CodeGenerator + ?Sized>(
let list_elem_ty = src_lst.get_type().element_type();
match list_elem_ty {
AnyTypeEnum::PointerType(ptr_ty) if ListType::is_type(ptr_ty, llvm_usize).is_ok() => {
AnyTypeEnum::PointerType(ptr_ty)
if ListType::is_representable(ptr_ty, llvm_usize).is_ok() =>
{
// The stride of elements in this dimension, i.e. the number of elements between arr[i]
// and arr[i + 1] in this dimension
let stride = call_ndarray_calc_size(
generator,
ctx,
&dst_arr.dim_sizes(),
&dst_arr.shape(),
(Some(llvm_usize.const_int(dim + 1, false)), None),
);
@ -717,11 +663,25 @@ fn ndarray_from_ndlist_impl<'ctx, G: CodeGenerator + ?Sized>(
|_, _| Ok(llvm_usize.const_int(1, false)),
|generator, ctx, _, i| {
let offset = ctx.builder.build_int_mul(stride, i, "").unwrap();
let offset = ctx
.builder
.build_int_mul(
offset,
ctx.builder
.build_int_truncate_or_bit_cast(
dst_arr.get_type().element_type().size_of().unwrap(),
offset.get_type(),
"",
)
.unwrap(),
"",
)
.unwrap();
let dst_ptr =
unsafe { ctx.builder.build_gep(dst_slice_ptr, &[offset], "").unwrap() };
let nested_lst_elem = ListValue::from_ptr_val(
let nested_lst_elem = ListValue::from_pointer_value(
unsafe { src_lst.data().get_unchecked(ctx, generator, &i, None) }
.into_pointer_value(),
llvm_usize,
@ -731,7 +691,6 @@ fn ndarray_from_ndlist_impl<'ctx, G: CodeGenerator + ?Sized>(
ndarray_from_ndlist_impl(
generator,
ctx,
elem_ty,
(dst_arr, dst_ptr),
nested_lst_elem,
dim + 1,
@ -742,14 +701,17 @@ fn ndarray_from_ndlist_impl<'ctx, G: CodeGenerator + ?Sized>(
)?;
}
AnyTypeEnum::PointerType(ptr_ty) if NDArrayType::is_type(ptr_ty, llvm_usize).is_ok() => {
AnyTypeEnum::PointerType(ptr_ty)
if NDArrayType::is_representable(ptr_ty, llvm_usize).is_ok() =>
{
todo!("Not implemented for list[ndarray]")
}
_ => {
let lst_len = src_lst.load_size(ctx, None);
let sizeof_elem = ctx.get_llvm_type(generator, elem_ty).size_of().unwrap();
let sizeof_elem = ctx.builder.build_int_cast(sizeof_elem, llvm_usize, "").unwrap();
let sizeof_elem = dst_arr.get_type().element_type().size_of().unwrap();
let sizeof_elem =
ctx.builder.build_int_z_extend_or_bit_cast(sizeof_elem, llvm_usize, "").unwrap();
let cpy_len = ctx
.builder
@ -803,8 +765,9 @@ fn call_ndarray_array_impl<'ctx, G: CodeGenerator + ?Sized>(
let object = object.into_pointer_value();
// object is an NDArray instance - copy object unless copy=0 && ndmin < object.ndims
if NDArrayValue::is_instance(object, llvm_usize).is_ok() {
let object = NDArrayValue::from_ptr_val(object, llvm_usize, None);
if NDArrayValue::is_representable(object, llvm_usize).is_ok() {
let llvm_elem_ty = ctx.get_llvm_type(generator, elem_ty);
let object = NDArrayValue::from_pointer_value(object, llvm_elem_ty, None, llvm_usize, None);
let ndarray = gen_if_else_expr_callback(
generator,
@ -866,7 +829,6 @@ fn call_ndarray_array_impl<'ctx, G: CodeGenerator + ?Sized>(
ndarray_sliced_copyto_impl(
generator,
ctx,
elem_ty,
(ndarray, ndarray.data().base_ptr(ctx, generator)),
(object, object.data().base_ptr(ctx, generator)),
0,
@ -878,16 +840,18 @@ fn call_ndarray_array_impl<'ctx, G: CodeGenerator + ?Sized>(
|_, _| Ok(Some(object.as_base_value())),
)?;
return Ok(NDArrayValue::from_ptr_val(
return Ok(NDArrayValue::from_pointer_value(
ndarray.map(BasicValueEnum::into_pointer_value).unwrap(),
llvm_elem_ty,
None,
llvm_usize,
None,
));
}
// Remaining case: TList
assert!(ListValue::is_instance(object, llvm_usize).is_ok());
let object = ListValue::from_ptr_val(object, llvm_usize, None);
assert!(ListValue::is_representable(object, llvm_usize).is_ok());
let object = ListValue::from_pointer_value(object, llvm_usize, None);
// The number of dimensions to prepend 1's to
let ndims = llvm_ndlist_get_ndims(generator, ctx, object.as_base_value().get_type());
@ -942,7 +906,7 @@ fn call_ndarray_array_impl<'ctx, G: CodeGenerator + ?Sized>(
.build_store(
lst,
ctx.builder
.build_bitcast(object.as_base_value(), llvm_plist_i8, "")
.build_bit_cast(object.as_base_value(), llvm_plist_i8, "")
.unwrap(),
)
.unwrap();
@ -964,10 +928,11 @@ fn call_ndarray_array_impl<'ctx, G: CodeGenerator + ?Sized>(
.builder
.build_load(lst, "")
.map(BasicValueEnum::into_pointer_value)
.map(|v| ctx.builder.build_bitcast(v, plist_plist_i8, "").unwrap())
.map(|v| ctx.builder.build_bit_cast(v, plist_plist_i8, "").unwrap())
.map(BasicValueEnum::into_pointer_value)
.unwrap();
let this_dim = ListValue::from_ptr_val(this_dim, llvm_usize, None);
let this_dim =
ListValue::from_pointer_value(this_dim, llvm_usize, None);
// TODO: Assert this_dim.sz != 0
@ -983,7 +948,9 @@ fn call_ndarray_array_impl<'ctx, G: CodeGenerator + ?Sized>(
ctx.builder
.build_store(
lst,
ctx.builder.build_bitcast(next_dim, llvm_plist_i8, "").unwrap(),
ctx.builder
.build_bit_cast(next_dim, llvm_plist_i8, "")
.unwrap(),
)
.unwrap();
@ -991,7 +958,7 @@ fn call_ndarray_array_impl<'ctx, G: CodeGenerator + ?Sized>(
},
)?;
let lst = ListValue::from_ptr_val(
let lst = ListValue::from_pointer_value(
ctx.builder
.build_load(lst, "")
.map(BasicValueEnum::into_pointer_value)
@ -1011,7 +978,6 @@ fn call_ndarray_array_impl<'ctx, G: CodeGenerator + ?Sized>(
ndarray_from_ndlist_impl(
generator,
ctx,
elem_ty,
(ndarray, ndarray.data().base_ptr(ctx, generator)),
object,
0,
@ -1071,7 +1037,7 @@ fn call_ndarray_eye_impl<'ctx, G: CodeGenerator + ?Sized>(
/// Copies a slice of an [`NDArrayValue`] to another.
///
/// - `dst_arr`: The [`NDArrayValue`] instance of the destination array. The `ndims` and `dim_sz`
/// - `dst_arr`: The [`NDArrayValue`] instance of the destination array. The `ndims` and `shape`
/// fields should be populated before calling this function.
/// - `dst_slice_ptr`: The [`PointerValue`] to the first element of the currently processing
/// dimensional slice in the destination array.
@ -1084,7 +1050,6 @@ fn call_ndarray_eye_impl<'ctx, G: CodeGenerator + ?Sized>(
fn ndarray_sliced_copyto_impl<'ctx, G: CodeGenerator + ?Sized>(
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
elem_ty: Type,
(dst_arr, dst_slice_ptr): (NDArrayValue<'ctx>, PointerValue<'ctx>),
(src_arr, src_slice_ptr): (NDArrayValue<'ctx>, PointerValue<'ctx>),
dim: u64,
@ -1093,14 +1058,16 @@ fn ndarray_sliced_copyto_impl<'ctx, G: CodeGenerator + ?Sized>(
let llvm_i1 = ctx.ctx.bool_type();
let llvm_usize = generator.get_size_type(ctx.ctx);
assert_eq!(dst_arr.get_type().element_type(), src_arr.get_type().element_type());
let sizeof_elem = dst_arr.get_type().element_type().size_of().unwrap();
// If there are no (remaining) slice expressions, memcpy the entire dimension
if slices.is_empty() {
let sizeof_elem = ctx.get_llvm_type(generator, elem_ty).size_of().unwrap();
let stride = call_ndarray_calc_size(
generator,
ctx,
&src_arr.dim_sizes(),
&src_arr.shape(),
(Some(llvm_usize.const_int(dim, false)), None),
);
let stride =
@ -1118,13 +1085,13 @@ fn ndarray_sliced_copyto_impl<'ctx, G: CodeGenerator + ?Sized>(
let src_stride = call_ndarray_calc_size(
generator,
ctx,
&src_arr.dim_sizes(),
&src_arr.shape(),
(Some(llvm_usize.const_int(dim + 1, false)), None),
);
let dst_stride = call_ndarray_calc_size(
generator,
ctx,
&dst_arr.dim_sizes(),
&dst_arr.shape(),
(Some(llvm_usize.const_int(dim + 1, false)), None),
);
@ -1147,9 +1114,29 @@ fn ndarray_sliced_copyto_impl<'ctx, G: CodeGenerator + ?Sized>(
|generator, ctx, _, src_i| {
// Calculate the offset of the active slice
let src_data_offset = ctx.builder.build_int_mul(src_stride, src_i, "").unwrap();
let src_data_offset = ctx
.builder
.build_int_mul(
src_data_offset,
ctx.builder
.build_int_z_extend_or_bit_cast(sizeof_elem, src_data_offset.get_type(), "")
.unwrap(),
"",
)
.unwrap();
let dst_i =
ctx.builder.build_load(dst_i_addr, "").map(BasicValueEnum::into_int_value).unwrap();
let dst_data_offset = ctx.builder.build_int_mul(dst_stride, dst_i, "").unwrap();
let dst_data_offset = ctx
.builder
.build_int_mul(
dst_data_offset,
ctx.builder
.build_int_z_extend_or_bit_cast(sizeof_elem, dst_data_offset.get_type(), "")
.unwrap(),
"",
)
.unwrap();
let (src_ptr, dst_ptr) = unsafe {
(
@ -1161,7 +1148,6 @@ fn ndarray_sliced_copyto_impl<'ctx, G: CodeGenerator + ?Sized>(
ndarray_sliced_copyto_impl(
generator,
ctx,
elem_ty,
(dst_arr, dst_ptr),
(src_arr, src_ptr),
dim + 1,
@ -1195,8 +1181,10 @@ pub fn ndarray_sliced_copy<'ctx, G: CodeGenerator + ?Sized>(
) -> Result<NDArrayValue<'ctx>, String> {
let llvm_i32 = ctx.ctx.i32_type();
let llvm_usize = generator.get_size_type(ctx.ctx);
let llvm_elem_ty = ctx.get_llvm_type(generator, elem_ty);
let ndarray = if slices.is_empty() {
let ndarray =
if slices.is_empty() {
create_ndarray_dyn_shape(
generator,
ctx,
@ -1204,15 +1192,12 @@ pub fn ndarray_sliced_copy<'ctx, G: CodeGenerator + ?Sized>(
&this,
|_, ctx, shape| Ok(shape.load_ndims(ctx)),
|generator, ctx, shape, idx| unsafe {
Ok(shape.dim_sizes().get_typed_unchecked(ctx, generator, &idx, None))
Ok(shape.shape().get_typed_unchecked(ctx, generator, &idx, None))
},
)?
} else {
let ndarray = create_ndarray_uninitialized(generator, ctx, elem_ty)?;
ndarray.store_ndims(ctx, generator, this.load_ndims(ctx));
let ndims = this.load_ndims(ctx);
ndarray.create_dim_sizes(ctx, llvm_usize, ndims);
let ndarray = NDArrayType::new(generator, ctx.ctx, llvm_elem_ty, None)
.construct_dyn_ndims(generator, ctx, this.load_ndims(ctx), None);
// Populate the first slices.len() dimensions by computing the size of each dim slice
for (i, (start, stop, step)) in slices.iter().enumerate() {
@ -1244,7 +1229,7 @@ pub fn ndarray_sliced_copy<'ctx, G: CodeGenerator + ?Sized>(
ctx.builder.build_int_z_extend_or_bit_cast(slice_len, llvm_usize, "").unwrap();
unsafe {
ndarray.dim_sizes().set_typed_unchecked(
ndarray.shape().set_typed_unchecked(
ctx,
generator,
&llvm_usize.const_int(i as u64, false),
@ -1262,8 +1247,8 @@ pub fn ndarray_sliced_copy<'ctx, G: CodeGenerator + ?Sized>(
(this.load_ndims(ctx), false),
|generator, ctx, _, idx| {
unsafe {
let dim_sz = this.dim_sizes().get_typed_unchecked(ctx, generator, &idx, None);
ndarray.dim_sizes().set_typed_unchecked(ctx, generator, &idx, dim_sz);
let shape = this.shape().get_typed_unchecked(ctx, generator, &idx, None);
ndarray.shape().set_typed_unchecked(ctx, generator, &idx, shape);
}
Ok(())
@ -1272,13 +1257,14 @@ pub fn ndarray_sliced_copy<'ctx, G: CodeGenerator + ?Sized>(
)
.unwrap();
ndarray_init_data(generator, ctx, elem_ty, ndarray)
unsafe { ndarray.create_data(generator, ctx) };
ndarray
};
ndarray_sliced_copyto_impl(
generator,
ctx,
elem_ty,
(ndarray, ndarray.data().base_ptr(ctx, generator)),
(this, this.data().base_ptr(ctx, generator)),
0,
@ -1324,7 +1310,7 @@ where
&operand,
|_, ctx, v| Ok(v.load_ndims(ctx)),
|generator, ctx, v, idx| unsafe {
Ok(v.dim_sizes().get_typed_unchecked(ctx, generator, &idx, None))
Ok(v.shape().get_typed_unchecked(ctx, generator, &idx, None))
},
)
.unwrap()
@ -1361,8 +1347,8 @@ pub fn ndarray_elementwise_binop_impl<'ctx, 'a, G, ValueFn>(
ctx: &mut CodeGenContext<'ctx, 'a>,
elem_ty: Type,
res: Option<NDArrayValue<'ctx>>,
lhs: (BasicValueEnum<'ctx>, bool),
rhs: (BasicValueEnum<'ctx>, bool),
lhs: (Type, BasicValueEnum<'ctx>, bool),
rhs: (Type, BasicValueEnum<'ctx>, bool),
value_fn: ValueFn,
) -> Result<NDArrayValue<'ctx>, String>
where
@ -1373,10 +1359,8 @@ where
(BasicValueEnum<'ctx>, BasicValueEnum<'ctx>),
) -> Result<BasicValueEnum<'ctx>, String>,
{
let llvm_usize = generator.get_size_type(ctx.ctx);
let (lhs_val, lhs_scalar) = lhs;
let (rhs_val, rhs_scalar) = rhs;
let (lhs_ty, lhs_val, lhs_scalar) = lhs;
let (rhs_ty, rhs_val, rhs_scalar) = rhs;
assert!(
!(lhs_scalar && rhs_scalar),
@ -1387,10 +1371,10 @@ where
let ndarray = res.unwrap_or_else(|| {
if lhs_scalar && rhs_scalar {
let lhs_val =
NDArrayValue::from_ptr_val(lhs_val.into_pointer_value(), llvm_usize, None);
let rhs_val =
NDArrayValue::from_ptr_val(rhs_val.into_pointer_value(), llvm_usize, None);
let lhs_val = NDArrayType::from_unifier_type(generator, ctx, lhs_ty)
.map_value(lhs_val.into_pointer_value(), None);
let rhs_val = NDArrayType::from_unifier_type(generator, ctx, rhs_ty)
.map_value(rhs_val.into_pointer_value(), None);
let ndarray_dims = call_ndarray_calc_broadcast(generator, ctx, lhs_val, rhs_val);
@ -1406,11 +1390,12 @@ where
)
.unwrap()
} else {
let ndarray = NDArrayValue::from_ptr_val(
if lhs_scalar { rhs_val } else { lhs_val }.into_pointer_value(),
llvm_usize,
None,
);
let ndarray = NDArrayType::from_unifier_type(
generator,
ctx,
if lhs_scalar { rhs_ty } else { lhs_ty },
)
.map_value(if lhs_scalar { rhs_val } else { lhs_val }.into_pointer_value(), None);
create_ndarray_dyn_shape(
generator,
@ -1419,7 +1404,7 @@ where
&ndarray,
|_, ctx, v| Ok(v.load_ndims(ctx)),
|generator, ctx, v, idx| unsafe {
Ok(v.dim_sizes().get_typed_unchecked(ctx, generator, &idx, None))
Ok(v.shape().get_typed_unchecked(ctx, generator, &idx, None))
},
)
.unwrap()
@ -1480,10 +1465,10 @@ pub fn ndarray_matmul_2d<'ctx, G: CodeGenerator>(
if let Some(res) = res {
let res_ndims = res.load_ndims(ctx);
let res_dim0 = unsafe {
res.dim_sizes().get_typed_unchecked(ctx, generator, &llvm_usize.const_zero(), None)
res.shape().get_typed_unchecked(ctx, generator, &llvm_usize.const_zero(), None)
};
let res_dim1 = unsafe {
res.dim_sizes().get_typed_unchecked(
res.shape().get_typed_unchecked(
ctx,
generator,
&llvm_usize.const_int(1, false),
@ -1491,10 +1476,10 @@ pub fn ndarray_matmul_2d<'ctx, G: CodeGenerator>(
)
};
let lhs_dim0 = unsafe {
lhs.dim_sizes().get_typed_unchecked(ctx, generator, &llvm_usize.const_zero(), None)
lhs.shape().get_typed_unchecked(ctx, generator, &llvm_usize.const_zero(), None)
};
let rhs_dim1 = unsafe {
rhs.dim_sizes().get_typed_unchecked(
rhs.shape().get_typed_unchecked(
ctx,
generator,
&llvm_usize.const_int(1, false),
@ -1543,15 +1528,10 @@ pub fn ndarray_matmul_2d<'ctx, G: CodeGenerator>(
if ctx.registry.llvm_options.opt_level == OptimizationLevel::None {
let lhs_dim1 = unsafe {
lhs.dim_sizes().get_typed_unchecked(
ctx,
generator,
&llvm_usize.const_int(1, false),
None,
)
lhs.shape().get_typed_unchecked(ctx, generator, &llvm_usize.const_int(1, false), None)
};
let rhs_dim0 = unsafe {
rhs.dim_sizes().get_typed_unchecked(ctx, generator, &llvm_usize.const_zero(), None)
rhs.shape().get_typed_unchecked(ctx, generator, &llvm_usize.const_zero(), None)
};
// lhs.dims[1] == rhs.dims[0]
@ -1590,7 +1570,7 @@ pub fn ndarray_matmul_2d<'ctx, G: CodeGenerator>(
},
|generator, ctx| {
Ok(Some(unsafe {
lhs.dim_sizes().get_typed_unchecked(
lhs.shape().get_typed_unchecked(
ctx,
generator,
&llvm_usize.const_zero(),
@ -1600,7 +1580,7 @@ pub fn ndarray_matmul_2d<'ctx, G: CodeGenerator>(
},
|generator, ctx| {
Ok(Some(unsafe {
rhs.dim_sizes().get_typed_unchecked(
rhs.shape().get_typed_unchecked(
ctx,
generator,
&llvm_usize.const_int(1, false),
@ -1627,7 +1607,7 @@ pub fn ndarray_matmul_2d<'ctx, G: CodeGenerator>(
let common_dim = {
let lhs_idx1 = unsafe {
lhs.dim_sizes().get_typed_unchecked(
lhs.shape().get_typed_unchecked(
ctx,
generator,
&llvm_usize.const_int(1, false),
@ -1635,7 +1615,7 @@ pub fn ndarray_matmul_2d<'ctx, G: CodeGenerator>(
)
};
let rhs_idx0 = unsafe {
rhs.dim_sizes().get_typed_unchecked(ctx, generator, &llvm_usize.const_zero(), None)
rhs.shape().get_typed_unchecked(ctx, generator, &llvm_usize.const_zero(), None)
};
let idx = llvm_intrinsics::call_expect(ctx, rhs_idx0, lhs_idx1, None);
@ -1959,18 +1939,18 @@ pub fn gen_ndarray_copy<'ctx>(
assert!(obj.is_some());
assert!(args.is_empty());
let llvm_usize = generator.get_size_type(context.ctx);
let this_ty = obj.as_ref().unwrap().0;
let (this_elem_ty, _) = unpack_ndarray_var_tys(&mut context.unifier, this_ty);
let this_arg =
obj.as_ref().unwrap().1.clone().to_basic_value_enum(context, generator, this_ty)?;
let llvm_this_ty = NDArrayType::from_unifier_type(generator, context, this_ty);
ndarray_copy_impl(
generator,
context,
this_elem_ty,
NDArrayValue::from_ptr_val(this_arg.into_pointer_value(), llvm_usize, None),
llvm_this_ty.map_value(this_arg.into_pointer_value(), None),
)
.map(NDArrayValue::into)
}
@ -1986,8 +1966,6 @@ pub fn gen_ndarray_fill<'ctx>(
assert!(obj.is_some());
assert_eq!(args.len(), 1);
let llvm_usize = generator.get_size_type(context.ctx);
let this_ty = obj.as_ref().unwrap().0;
let this_arg = obj
.as_ref()
@ -1999,10 +1977,12 @@ pub fn gen_ndarray_fill<'ctx>(
let value_ty = fun.0.args[0].ty;
let value_arg = args[0].1.clone().to_basic_value_enum(context, generator, value_ty)?;
let llvm_this_ty = NDArrayType::from_unifier_type(generator, context, this_ty);
ndarray_fill_flattened(
generator,
context,
NDArrayValue::from_ptr_val(this_arg, llvm_usize, None),
llvm_this_ty.map_value(this_arg, None),
|generator, ctx, _| {
let value = if value_arg.is_pointer_value() {
let llvm_i1 = ctx.ctx.bool_type();
@ -2035,16 +2015,17 @@ pub fn gen_ndarray_fill<'ctx>(
pub fn ndarray_transpose<'ctx, G: CodeGenerator + ?Sized>(
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
x1: (Type, BasicValueEnum<'ctx>),
(x1_ty, x1): (Type, BasicValueEnum<'ctx>),
) -> Result<BasicValueEnum<'ctx>, String> {
const FN_NAME: &str = "ndarray_transpose";
let (x1_ty, x1) = x1;
let llvm_usize = generator.get_size_type(ctx.ctx);
if let BasicValueEnum::PointerValue(n1) = x1 {
let llvm_ndarray_ty = NDArrayType::from_unifier_type(generator, ctx, x1_ty);
let (elem_ty, _) = unpack_ndarray_var_tys(&mut ctx.unifier, x1_ty);
let n1 = NDArrayValue::from_ptr_val(n1, llvm_usize, None);
let n_sz = call_ndarray_calc_size(generator, ctx, &n1.dim_sizes(), (None, None));
let n1 = llvm_ndarray_ty.map_value(n1, None);
let n_sz = call_ndarray_calc_size(generator, ctx, &n1.shape(), (None, None));
// Dimensions are reversed in the transposed array
let out = create_ndarray_dyn_shape(
@ -2059,7 +2040,7 @@ pub fn ndarray_transpose<'ctx, G: CodeGenerator + ?Sized>(
.builder
.build_int_sub(new_idx, new_idx.get_type().const_int(1, false), "")
.unwrap();
unsafe { Ok(n.dim_sizes().get_typed_unchecked(ctx, generator, &new_idx, None)) }
unsafe { Ok(n.shape().get_typed_unchecked(ctx, generator, &new_idx, None)) }
},
)
.unwrap();
@ -2096,7 +2077,7 @@ pub fn ndarray_transpose<'ctx, G: CodeGenerator + ?Sized>(
.build_int_sub(ndim_rev, llvm_usize.const_int(1, false), "")
.unwrap();
let dim = unsafe {
n1.dim_sizes().get_typed_unchecked(ctx, generator, &ndim_rev, None)
n1.shape().get_typed_unchecked(ctx, generator, &ndim_rev, None)
};
let rem_idx_val =
@ -2162,8 +2143,9 @@ pub fn ndarray_reshape<'ctx, G: CodeGenerator + ?Sized>(
if let BasicValueEnum::PointerValue(n1) = x1 {
let (elem_ty, _) = unpack_ndarray_var_tys(&mut ctx.unifier, x1_ty);
let n1 = NDArrayValue::from_ptr_val(n1, llvm_usize, None);
let n_sz = call_ndarray_calc_size(generator, ctx, &n1.dim_sizes(), (None, None));
let llvm_ndarray_ty = NDArrayType::from_unifier_type(generator, ctx, x1_ty);
let n1 = llvm_ndarray_ty.map_value(n1, None);
let n_sz = call_ndarray_calc_size(generator, ctx, &n1.shape(), (None, None));
let acc = generator.gen_var_alloc(ctx, llvm_usize.into(), None)?;
let num_neg = generator.gen_var_alloc(ctx, llvm_usize.into(), None)?;
@ -2172,11 +2154,11 @@ pub fn ndarray_reshape<'ctx, G: CodeGenerator + ?Sized>(
let out = match shape {
BasicValueEnum::PointerValue(shape_list_ptr)
if ListValue::is_instance(shape_list_ptr, llvm_usize).is_ok() =>
if ListValue::is_representable(shape_list_ptr, llvm_usize).is_ok() =>
{
// 1. A list of ints; e.g., `np.reshape(arr, [int64(600), int64(800, -1])`
let shape_list = ListValue::from_ptr_val(shape_list_ptr, llvm_usize, None);
let shape_list = ListValue::from_pointer_value(shape_list_ptr, llvm_usize, None);
// Check for -1 in dimensions
gen_for_callback_incrementing(
generator,
@ -2391,7 +2373,7 @@ pub fn ndarray_reshape<'ctx, G: CodeGenerator + ?Sized>(
);
// The new shape must be compatible with the old shape
let out_sz = call_ndarray_calc_size(generator, ctx, &out.dim_sizes(), (None, None));
let out_sz = call_ndarray_calc_size(generator, ctx, &out.shape(), (None, None));
ctx.make_assert(
generator,
ctx.builder.build_int_compare(IntPredicate::EQ, out_sz, n_sz, "").unwrap(),
@ -2439,17 +2421,17 @@ pub fn ndarray_dot<'ctx, G: CodeGenerator + ?Sized>(
) -> Result<BasicValueEnum<'ctx>, String> {
const FN_NAME: &str = "ndarray_dot";
let (x1_ty, x1) = x1;
let (_, x2) = x2;
let (x2_ty, x2) = x2;
let llvm_usize = generator.get_size_type(ctx.ctx);
match (x1, x2) {
(BasicValueEnum::PointerValue(n1), BasicValueEnum::PointerValue(n2)) => {
let n1 = NDArrayValue::from_ptr_val(n1, llvm_usize, None);
let n2 = NDArrayValue::from_ptr_val(n2, llvm_usize, None);
let n1 = NDArrayType::from_unifier_type(generator, ctx, x1_ty).map_value(n1, None);
let n2 = NDArrayType::from_unifier_type(generator, ctx, x2_ty).map_value(n2, None);
let n1_sz = call_ndarray_calc_size(generator, ctx, &n1.dim_sizes(), (None, None));
let n2_sz = call_ndarray_calc_size(generator, ctx, &n1.dim_sizes(), (None, None));
let n1_sz = call_ndarray_calc_size(generator, ctx, &n1.shape(), (None, None));
let n2_sz = call_ndarray_calc_size(generator, ctx, &n1.shape(), (None, None));
ctx.make_assert(
generator,
@ -2486,7 +2468,7 @@ pub fn ndarray_dot<'ctx, G: CodeGenerator + ?Sized>(
.build_float_mul(e1, elem2.into_float_value(), "")
.unwrap()
.as_basic_value_enum(),
_ => codegen_unreachable!(ctx),
_ => codegen_unreachable!(ctx, "product: {}", elem1.get_type()),
};
let acc_val = ctx.builder.build_load(acc, "").unwrap();
let acc_val = match acc_val {
@ -2500,7 +2482,7 @@ pub fn ndarray_dot<'ctx, G: CodeGenerator + ?Sized>(
.build_float_add(e1, product.into_float_value(), "")
.unwrap()
.as_basic_value_enum(),
_ => codegen_unreachable!(ctx),
_ => codegen_unreachable!(ctx, "acc_val: {}", acc_val.get_type()),
};
ctx.builder.build_store(acc, acc_val).unwrap();

View File

@ -1,9 +1,22 @@
use inkwell::{
attributes::{Attribute, AttributeLoc},
basic_block::BasicBlock,
types::{BasicType, BasicTypeEnum},
values::{BasicValue, BasicValueEnum, FunctionValue, IntValue, PointerValue},
IntPredicate,
};
use itertools::{izip, Itertools};
use nac3parser::ast::{
Constant, ExcepthandlerKind, Expr, ExprKind, Location, Stmt, StmtKind, StrRef,
};
use super::{
classes::{ArrayLikeIndexer, ArraySliceValue, ListValue, RangeValue},
expr::{destructure_range, gen_binop_expr},
gen_in_range_check,
irrt::{handle_slice_indices, list_slice_assignment},
macros::codegen_unreachable,
values::{ArrayLikeIndexer, ArraySliceValue, ListValue, RangeValue},
CodeGenContext, CodeGenerator,
};
use crate::{
@ -14,17 +27,6 @@ use crate::{
typedef::{iter_type_vars, FunSignature, Type, TypeEnum},
},
};
use inkwell::{
attributes::{Attribute, AttributeLoc},
basic_block::BasicBlock,
types::{BasicType, BasicTypeEnum},
values::{BasicValue, BasicValueEnum, FunctionValue, IntValue, PointerValue},
IntPredicate,
};
use itertools::{izip, Itertools};
use nac3parser::ast::{
Constant, ExcepthandlerKind, Expr, ExprKind, Location, Stmt, StmtKind, StrRef,
};
/// See [`CodeGenerator::gen_var_alloc`].
pub fn gen_var<'ctx>(
@ -174,6 +176,14 @@ pub fn gen_assign<'ctx, G: CodeGenerator>(
}
}
let val = value.to_basic_value_enum(ctx, generator, target.custom.unwrap())?;
// Perform i1 <-> i8 conversion as needed
let val = if ctx.unifier.unioned(target.custom.unwrap(), ctx.primitives.bool) {
generator.bool_to_i8(ctx, val.into_int_value()).into()
} else {
val
};
ctx.builder.build_store(ptr, val).unwrap();
}
};
@ -300,7 +310,7 @@ pub fn gen_setitem<'ctx, G: CodeGenerator>(
.unwrap()
.to_basic_value_enum(ctx, generator, target_ty)?
.into_pointer_value();
let target = ListValue::from_ptr_val(target, llvm_usize, None);
let target = ListValue::from_pointer_value(target, llvm_usize, None);
if let ExprKind::Slice { .. } = &key.node {
// Handle assigning to a slice
@ -321,7 +331,7 @@ pub fn gen_setitem<'ctx, G: CodeGenerator>(
let value =
value.to_basic_value_enum(ctx, generator, value_ty)?.into_pointer_value();
let value = ListValue::from_ptr_val(value, llvm_usize, None);
let value = ListValue::from_pointer_value(value, llvm_usize, None);
let target_item_ty = ctx.get_llvm_type(generator, target_item_ty);
let Some(src_ind) = handle_slice_indices(
@ -453,7 +463,8 @@ pub fn gen_for<G: CodeGenerator>(
TypeEnum::TObj { obj_id, .. }
if *obj_id == ctx.primitives.range.obj_id(&ctx.unifier).unwrap() =>
{
let iter_val = RangeValue::from_ptr_val(iter_val.into_pointer_value(), Some("range"));
let iter_val =
RangeValue::from_pointer_value(iter_val.into_pointer_value(), Some("range"));
// Internal variable for loop; Cannot be assigned
let i = generator.gen_var_alloc(ctx, int32.into(), Some("for.i.addr"))?;
// Variable declared in "target" expression of the loop; Can be reassigned *or* shadowed
@ -1665,6 +1676,23 @@ pub fn gen_return<G: CodeGenerator>(
} else {
None
};
// Remap boolean return type into i1
let value = value.map(|ret_val| {
// The "return type" of a sret function is in the first parameter
let expected_ty = if ctx.need_sret {
func.get_type().get_param_types()[0]
} else {
func.get_type().get_return_type().unwrap()
};
if matches!(expected_ty, BasicTypeEnum::IntType(ty) if ty.get_bit_width() == 1) {
generator.bool_to_i1(ctx, ret_val.into_int_value()).into()
} else {
ret_val
}
});
if let Some(return_target) = ctx.return_target {
if let Some(value) = value {
ctx.builder.build_store(ctx.return_buffer.unwrap(), value).unwrap();
@ -1675,25 +1703,6 @@ pub fn gen_return<G: CodeGenerator>(
ctx.builder.build_store(ctx.return_buffer.unwrap(), value.unwrap()).unwrap();
ctx.builder.build_return(None).unwrap();
} else {
// Remap boolean return type into i1
let value = value.map(|v| {
let expected_ty = func.get_type().get_return_type().unwrap();
let ret_val = v.as_basic_value_enum();
if expected_ty.is_int_type() && ret_val.is_int_value() {
let ret_type = expected_ty.into_int_type();
let ret_val = ret_val.into_int_value();
if ret_type.get_bit_width() == 1 && ret_val.get_type().get_bit_width() != 1 {
generator.bool_to_i1(ctx, ret_val)
} else {
ret_val
}
.into()
} else {
ret_val
}
});
let value = value.as_ref().map(|v| v as &dyn BasicValue);
ctx.builder.build_return(value).unwrap();
}
@ -1820,6 +1829,37 @@ pub fn gen_stmt<G: CodeGenerator>(
stmt.location,
);
}
StmtKind::Global { names, .. } => {
let registered_globals = ctx
.top_level
.definitions
.read()
.iter()
.filter_map(|def| {
if let TopLevelDef::Variable { simple_name, ty, .. } = &*def.read() {
Some((*simple_name, *ty))
} else {
None
}
})
.collect_vec();
for id in names {
let Some((_, ty)) = registered_globals.iter().find(|(name, _)| name == id) else {
return Err(format!("{id} is not a global at {}", stmt.location));
};
let resolver = ctx.resolver.clone();
let ptr = resolver
.get_symbol_value(*id, ctx, generator)
.map(|val| val.to_basic_value_enum(ctx, generator, *ty))
.transpose()?
.map(BasicValueEnum::into_pointer_value)
.unwrap();
ctx.var_assignment.insert(*id, (ptr, None, 0));
}
}
_ => unimplemented!(),
};
Ok(())

View File

@ -1,34 +1,37 @@
use crate::{
codegen::{
classes::{ListType, NDArrayType, ProxyType, RangeType},
concrete_type::ConcreteTypeStore,
CodeGenContext, CodeGenLLVMOptions, CodeGenTargetMachineOptions, CodeGenTask,
CodeGenerator, DefaultCodeGenerator, WithCall, WorkerRegistry,
},
symbol_resolver::{SymbolResolver, ValueEnum},
toplevel::{
composer::{ComposerConfig, TopLevelComposer},
DefinitionId, FunInstance, TopLevelContext, TopLevelDef,
},
typecheck::{
type_inferencer::{FunctionData, Inferencer, PrimitiveStore},
typedef::{FunSignature, FuncArg, Type, TypeEnum, Unifier, VarMap},
},
use std::{
collections::{HashMap, HashSet},
sync::Arc,
};
use indexmap::IndexMap;
use indoc::indoc;
use inkwell::{
targets::{InitializationConfig, Target},
OptimizationLevel,
};
use nac3parser::ast::FileName;
use nac3parser::{
ast::{fold::Fold, StrRef},
ast::{fold::Fold, FileName, StrRef},
parser::parse_program,
};
use parking_lot::RwLock;
use std::collections::{HashMap, HashSet};
use std::sync::Arc;
use super::{
concrete_type::ConcreteTypeStore,
types::{ndarray::NDArrayType, ListType, ProxyType, RangeType},
CodeGenContext, CodeGenLLVMOptions, CodeGenTargetMachineOptions, CodeGenTask, CodeGenerator,
DefaultCodeGenerator, WithCall, WorkerRegistry,
};
use crate::{
symbol_resolver::{SymbolResolver, ValueEnum},
toplevel::{
composer::{ComposerConfig, TopLevelComposer},
DefinitionId, FunInstance, TopLevelContext, TopLevelDef,
},
typecheck::{
type_inferencer::{FunctionData, IdentifierInfo, Inferencer, PrimitiveStore},
typedef::{FunSignature, FuncArg, Type, TypeEnum, Unifier, VarMap},
},
};
struct Resolver {
id_to_type: HashMap<StrRef, Type>,
@ -64,6 +67,7 @@ impl SymbolResolver for Resolver {
&self,
_: StrRef,
_: &mut CodeGenContext<'ctx, '_>,
_: &mut dyn CodeGenerator,
) -> Option<ValueEnum<'ctx>> {
unimplemented!()
}
@ -138,7 +142,8 @@ fn test_primitives() {
};
let mut virtual_checks = Vec::new();
let mut calls = HashMap::new();
let mut identifiers: HashSet<_> = ["a".into(), "b".into()].into();
let mut identifiers: HashMap<_, _> =
["a".into(), "b".into()].map(|id| (id, IdentifierInfo::default())).into();
let mut inferencer = Inferencer {
top_level: &top_level,
function_data: &mut function_data,
@ -317,7 +322,8 @@ fn test_simple_call() {
};
let mut virtual_checks = Vec::new();
let mut calls = HashMap::new();
let mut identifiers: HashSet<_> = ["a".into(), "foo".into()].into();
let mut identifiers: HashMap<_, _> =
["a".into(), "foo".into()].map(|id| (id, IdentifierInfo::default())).into();
let mut inferencer = Inferencer {
top_level: &top_level,
function_data: &mut function_data,
@ -446,7 +452,7 @@ fn test_classes_list_type_new() {
let llvm_usize = generator.get_size_type(&ctx);
let llvm_list = ListType::new(&generator, &ctx, llvm_i32.into());
assert!(ListType::is_type(llvm_list.as_base_type(), llvm_usize).is_ok());
assert!(ListType::is_representable(llvm_list.as_base_type(), llvm_usize).is_ok());
}
#[test]
@ -454,7 +460,7 @@ fn test_classes_range_type_new() {
let ctx = inkwell::context::Context::create();
let llvm_range = RangeType::new(&ctx);
assert!(RangeType::is_type(llvm_range.as_base_type()).is_ok());
assert!(RangeType::is_representable(llvm_range.as_base_type()).is_ok());
}
#[test]
@ -465,6 +471,6 @@ fn test_classes_ndarray_type_new() {
let llvm_i32 = ctx.i32_type();
let llvm_usize = generator.get_size_type(&ctx);
let llvm_ndarray = NDArrayType::new(&generator, &ctx, llvm_i32.into());
assert!(NDArrayType::is_type(llvm_ndarray.as_base_type(), llvm_usize).is_ok());
let llvm_ndarray = NDArrayType::new(&generator, &ctx, llvm_i32.into(), None);
assert!(NDArrayType::is_representable(llvm_ndarray.as_base_type(), llvm_usize).is_ok());
}

View File

@ -0,0 +1,206 @@
use inkwell::{
context::Context,
types::{AnyTypeEnum, BasicType, BasicTypeEnum, IntType, PointerType},
values::IntValue,
AddressSpace,
};
use super::ProxyType;
use crate::codegen::{
values::{ArraySliceValue, ListValue, ProxyValue},
CodeGenContext, CodeGenerator,
};
/// Proxy type for a `list` type in LLVM.
#[derive(Debug, PartialEq, Eq, Clone, Copy)]
pub struct ListType<'ctx> {
ty: PointerType<'ctx>,
llvm_usize: IntType<'ctx>,
}
impl<'ctx> ListType<'ctx> {
/// Checks whether `llvm_ty` represents a `list` type, returning [Err] if it does not.
pub fn is_representable(
llvm_ty: PointerType<'ctx>,
llvm_usize: IntType<'ctx>,
) -> Result<(), String> {
let llvm_list_ty = llvm_ty.get_element_type();
let AnyTypeEnum::StructType(llvm_list_ty) = llvm_list_ty else {
return Err(format!("Expected struct type for `list` type, got {llvm_list_ty}"));
};
if llvm_list_ty.count_fields() != 2 {
return Err(format!(
"Expected 2 fields in `list`, got {}",
llvm_list_ty.count_fields()
));
}
let list_size_ty = llvm_list_ty.get_field_type_at_index(0).unwrap();
let Ok(_) = PointerType::try_from(list_size_ty) else {
return Err(format!("Expected pointer type for `list.0`, got {list_size_ty}"));
};
let list_data_ty = llvm_list_ty.get_field_type_at_index(1).unwrap();
let Ok(list_data_ty) = IntType::try_from(list_data_ty) else {
return Err(format!("Expected int type for `list.1`, got {list_data_ty}"));
};
if list_data_ty.get_bit_width() != llvm_usize.get_bit_width() {
return Err(format!(
"Expected {}-bit int type for `list.1`, got {}-bit int",
llvm_usize.get_bit_width(),
list_data_ty.get_bit_width()
));
}
Ok(())
}
/// Creates an LLVM type corresponding to the expected structure of a `List`.
#[must_use]
fn llvm_type(
ctx: &'ctx Context,
element_type: BasicTypeEnum<'ctx>,
llvm_usize: IntType<'ctx>,
) -> PointerType<'ctx> {
// struct List { data: T*, size: size_t }
let field_tys = [element_type.ptr_type(AddressSpace::default()).into(), llvm_usize.into()];
ctx.struct_type(&field_tys, false).ptr_type(AddressSpace::default())
}
/// Creates an instance of [`ListType`].
#[must_use]
pub fn new<G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &'ctx Context,
element_type: BasicTypeEnum<'ctx>,
) -> Self {
let llvm_usize = generator.get_size_type(ctx);
let llvm_list = Self::llvm_type(ctx, element_type, llvm_usize);
ListType::from_type(llvm_list, llvm_usize)
}
/// Creates an [`ListType`] from a [`PointerType`].
#[must_use]
pub fn from_type(ptr_ty: PointerType<'ctx>, llvm_usize: IntType<'ctx>) -> Self {
debug_assert!(Self::is_representable(ptr_ty, llvm_usize).is_ok());
ListType { ty: ptr_ty, llvm_usize }
}
/// Returns the type of the `size` field of this `list` type.
#[must_use]
pub fn size_type(&self) -> IntType<'ctx> {
self.as_base_type()
.get_element_type()
.into_struct_type()
.get_field_type_at_index(1)
.map(BasicTypeEnum::into_int_type)
.unwrap()
}
/// Returns the element type of this `list` type.
#[must_use]
pub fn element_type(&self) -> AnyTypeEnum<'ctx> {
self.as_base_type()
.get_element_type()
.into_struct_type()
.get_field_type_at_index(0)
.map(BasicTypeEnum::into_pointer_type)
.map(PointerType::get_element_type)
.unwrap()
}
/// Allocates an instance of [`ListValue`] as if by calling `alloca` on the base type.
#[must_use]
pub fn alloca<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
name: Option<&'ctx str>,
) -> <Self as ProxyType<'ctx>>::Value {
<Self as ProxyType<'ctx>>::Value::from_pointer_value(
self.raw_alloca(generator, ctx, name),
self.llvm_usize,
name,
)
}
/// Converts an existing value into a [`ListValue`].
#[must_use]
pub fn map_value(
&self,
value: <<Self as ProxyType<'ctx>>::Value as ProxyValue<'ctx>>::Base,
name: Option<&'ctx str>,
) -> <Self as ProxyType<'ctx>>::Value {
<Self as ProxyType<'ctx>>::Value::from_pointer_value(value, self.llvm_usize, name)
}
}
impl<'ctx> ProxyType<'ctx> for ListType<'ctx> {
type Base = PointerType<'ctx>;
type Value = ListValue<'ctx>;
fn is_type<G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &'ctx Context,
llvm_ty: impl BasicType<'ctx>,
) -> Result<(), String> {
if let BasicTypeEnum::PointerType(ty) = llvm_ty.as_basic_type_enum() {
<Self as ProxyType<'ctx>>::is_representable(generator, ctx, ty)
} else {
Err(format!("Expected pointer type, got {llvm_ty:?}"))
}
}
fn is_representable<G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &'ctx Context,
llvm_ty: Self::Base,
) -> Result<(), String> {
Self::is_representable(llvm_ty, generator.get_size_type(ctx))
}
fn raw_alloca<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
name: Option<&'ctx str>,
) -> <Self::Value as ProxyValue<'ctx>>::Base {
generator
.gen_var_alloc(
ctx,
self.as_base_type().get_element_type().into_struct_type().into(),
name,
)
.unwrap()
}
fn array_alloca<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
size: IntValue<'ctx>,
name: Option<&'ctx str>,
) -> ArraySliceValue<'ctx> {
generator
.gen_array_var_alloc(
ctx,
self.as_base_type().get_element_type().into_struct_type().into(),
size,
name,
)
.unwrap()
}
fn as_base_type(&self) -> Self::Base {
self.ty
}
}
impl<'ctx> From<ListType<'ctx>> for PointerType<'ctx> {
fn from(value: ListType<'ctx>) -> Self {
value.as_base_type()
}
}

View File

@ -0,0 +1,76 @@
//! This module contains abstraction over all intrinsic composite types of NAC3.
//!
//! # `raw_alloca` vs `alloca` vs `construct`
//!
//! There are three ways of creating a new object instance using the abstractions provided by this
//! module.
//!
//! - `raw_alloca`: Allocates the object on the stack, returning an instance of
//! [`impl BasicValue`][inkwell::values::BasicValue]. This is similar to a `malloc` expression in
//! C++ but the object is allocated on the stack.
//! - `alloca`: Similar to `raw_alloca`, but also wraps the allocated object with
//! [`<Self as ProxyType<'ctx>>::Value`][ProxyValue], and returns the wrapped object. The returned
//! object will not initialize any value or fields. This is similar to a type-safe `malloc`
//! expression in C++ but the object is allocated on the stack.
//! - `construct`: Similar to `alloca`, but performs some initialization on the value or fields of
//! the returned object. This is similar to a `new` expression in C++ but the object is allocated
//! on the stack.
use inkwell::{context::Context, types::BasicType, values::IntValue};
use super::{
values::{ArraySliceValue, ProxyValue},
{CodeGenContext, CodeGenerator},
};
pub use list::*;
pub use range::*;
mod list;
pub mod ndarray;
mod range;
pub mod structure;
pub mod utils;
/// A LLVM type that is used to represent a corresponding type in NAC3.
pub trait ProxyType<'ctx>: Into<Self::Base> {
/// The LLVM type of which values of this type possess. This is usually a
/// [LLVM pointer type][PointerType] for any non-primitive types.
type Base: BasicType<'ctx>;
/// The type of values represented by this type.
type Value: ProxyValue<'ctx, Type = Self>;
fn is_type<G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &'ctx Context,
llvm_ty: impl BasicType<'ctx>,
) -> Result<(), String>;
/// Checks whether `llvm_ty` can be represented by this [`ProxyType`].
fn is_representable<G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &'ctx Context,
llvm_ty: Self::Base,
) -> Result<(), String>;
/// Creates a new value of this type, returning the LLVM instance of this value.
fn raw_alloca<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
name: Option<&'ctx str>,
) -> <Self::Value as ProxyValue<'ctx>>::Base;
/// Creates a new array value of this type, returning an [`ArraySliceValue`] encapsulating the
/// resulting array.
fn array_alloca<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
size: IntValue<'ctx>,
name: Option<&'ctx str>,
) -> ArraySliceValue<'ctx>;
/// Returns the [base type][Self::Base] of this proxy.
fn as_base_type(&self) -> Self::Base;
}

View File

@ -0,0 +1,257 @@
use inkwell::{
context::Context,
types::{AnyTypeEnum, BasicType, BasicTypeEnum, IntType, PointerType},
values::{IntValue, PointerValue},
AddressSpace,
};
use itertools::Itertools;
use nac3core_derive::StructFields;
use crate::{
codegen::{
types::{
structure::{
check_struct_type_matches_fields, FieldIndexCounter, StructField, StructFields,
},
ProxyType,
},
values::{ndarray::ContiguousNDArrayValue, ArraySliceValue, ProxyValue},
CodeGenContext, CodeGenerator,
},
toplevel::numpy::unpack_ndarray_var_tys,
typecheck::typedef::Type,
};
#[derive(Debug, PartialEq, Eq, Clone, Copy)]
pub struct ContiguousNDArrayType<'ctx> {
ty: PointerType<'ctx>,
item: BasicTypeEnum<'ctx>,
llvm_usize: IntType<'ctx>,
}
#[derive(PartialEq, Eq, Clone, Copy, StructFields)]
pub struct ContiguousNDArrayFields<'ctx> {
#[value_type(usize)]
pub ndims: StructField<'ctx, IntValue<'ctx>>,
#[value_type(usize.ptr_type(AddressSpace::default()))]
pub shape: StructField<'ctx, PointerValue<'ctx>>,
#[value_type(i8_type().ptr_type(AddressSpace::default()))]
pub data: StructField<'ctx, PointerValue<'ctx>>,
}
impl<'ctx> ContiguousNDArrayFields<'ctx> {
#[must_use]
pub fn new_typed(item: BasicTypeEnum<'ctx>, llvm_usize: IntType<'ctx>) -> Self {
let mut counter = FieldIndexCounter::default();
ContiguousNDArrayFields {
ndims: StructField::create(&mut counter, "ndims", llvm_usize),
shape: StructField::create(
&mut counter,
"shape",
llvm_usize.ptr_type(AddressSpace::default()),
),
data: StructField::create(&mut counter, "data", item.ptr_type(AddressSpace::default())),
}
}
}
impl<'ctx> ContiguousNDArrayType<'ctx> {
/// Checks whether `llvm_ty` represents a `ndarray` type, returning [Err] if it does not.
pub fn is_representable(
llvm_ty: PointerType<'ctx>,
llvm_usize: IntType<'ctx>,
) -> Result<(), String> {
let ctx = llvm_ty.get_context();
let llvm_ty = llvm_ty.get_element_type();
let AnyTypeEnum::StructType(llvm_ty) = llvm_ty else {
return Err(format!(
"Expected struct type for `ContiguousNDArray` type, got {llvm_ty}"
));
};
let fields = ContiguousNDArrayFields::new(ctx, llvm_usize);
check_struct_type_matches_fields(
fields,
llvm_ty,
"ContiguousNDArray",
&[(fields.data.name(), &|ty| {
if ty.is_pointer_type() {
Ok(())
} else {
Err(format!("Expected T* for `ContiguousNDArray.data`, got {ty}"))
}
})],
)
}
/// Returns an instance of [`StructFields`] containing all field accessors for this type.
#[must_use]
fn fields(
item: BasicTypeEnum<'ctx>,
llvm_usize: IntType<'ctx>,
) -> ContiguousNDArrayFields<'ctx> {
ContiguousNDArrayFields::new_typed(item, llvm_usize)
}
/// See [`NDArrayType::fields`].
// TODO: Move this into e.g. StructProxyType
#[must_use]
pub fn get_fields(&self) -> ContiguousNDArrayFields<'ctx> {
Self::fields(self.item, self.llvm_usize)
}
/// Creates an LLVM type corresponding to the expected structure of an `NDArray`.
#[must_use]
fn llvm_type(
ctx: &'ctx Context,
item: BasicTypeEnum<'ctx>,
llvm_usize: IntType<'ctx>,
) -> PointerType<'ctx> {
let field_tys =
Self::fields(item, llvm_usize).into_iter().map(|field| field.1).collect_vec();
ctx.struct_type(&field_tys, false).ptr_type(AddressSpace::default())
}
/// Creates an instance of [`ContiguousNDArrayType`].
#[must_use]
pub fn new<G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &'ctx Context,
item: BasicTypeEnum<'ctx>,
) -> Self {
let llvm_usize = generator.get_size_type(ctx);
let llvm_cndarray = Self::llvm_type(ctx, item, llvm_usize);
Self { ty: llvm_cndarray, item, llvm_usize }
}
/// Creates an [`ContiguousNDArrayType`] from a [unifier type][Type].
#[must_use]
pub fn from_unifier_type<G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &mut CodeGenContext<'ctx, '_>,
ty: Type,
) -> Self {
let (dtype, _) = unpack_ndarray_var_tys(&mut ctx.unifier, ty);
let llvm_dtype = ctx.get_llvm_type(generator, dtype);
let llvm_usize = generator.get_size_type(ctx.ctx);
Self { ty: Self::llvm_type(ctx.ctx, llvm_dtype, llvm_usize), item: llvm_dtype, llvm_usize }
}
/// Creates an [`ContiguousNDArrayType`] from a [`PointerType`] representing an `NDArray`.
#[must_use]
pub fn from_type(
ptr_ty: PointerType<'ctx>,
item: BasicTypeEnum<'ctx>,
llvm_usize: IntType<'ctx>,
) -> Self {
debug_assert!(Self::is_representable(ptr_ty, llvm_usize).is_ok());
Self { ty: ptr_ty, item, llvm_usize }
}
/// Allocates an instance of [`ContiguousNDArrayValue`] as if by calling `alloca` on the base type.
#[must_use]
pub fn alloca<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
name: Option<&'ctx str>,
) -> <Self as ProxyType<'ctx>>::Value {
<Self as ProxyType<'ctx>>::Value::from_pointer_value(
self.raw_alloca(generator, ctx, name),
self.item,
self.llvm_usize,
name,
)
}
/// Converts an existing value into a [`ContiguousNDArrayValue`].
#[must_use]
pub fn map_value(
&self,
value: <<Self as ProxyType<'ctx>>::Value as ProxyValue<'ctx>>::Base,
name: Option<&'ctx str>,
) -> <Self as ProxyType<'ctx>>::Value {
<Self as ProxyType<'ctx>>::Value::from_pointer_value(
value,
self.item,
self.llvm_usize,
name,
)
}
}
impl<'ctx> ProxyType<'ctx> for ContiguousNDArrayType<'ctx> {
type Base = PointerType<'ctx>;
type Value = ContiguousNDArrayValue<'ctx>;
fn is_type<G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &'ctx Context,
llvm_ty: impl BasicType<'ctx>,
) -> Result<(), String> {
if let BasicTypeEnum::PointerType(ty) = llvm_ty.as_basic_type_enum() {
<Self as ProxyType<'ctx>>::is_representable(generator, ctx, ty)
} else {
Err(format!("Expected pointer type, got {llvm_ty:?}"))
}
}
fn is_representable<G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &'ctx Context,
llvm_ty: Self::Base,
) -> Result<(), String> {
Self::is_representable(llvm_ty, generator.get_size_type(ctx))
}
fn raw_alloca<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
name: Option<&'ctx str>,
) -> <Self::Value as ProxyValue<'ctx>>::Base {
generator
.gen_var_alloc(
ctx,
self.as_base_type().get_element_type().into_struct_type().into(),
name,
)
.unwrap()
}
fn array_alloca<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
size: IntValue<'ctx>,
name: Option<&'ctx str>,
) -> ArraySliceValue<'ctx> {
generator
.gen_array_var_alloc(
ctx,
self.as_base_type().get_element_type().into_struct_type().into(),
size,
name,
)
.unwrap()
}
fn as_base_type(&self) -> Self::Base {
self.ty
}
}
impl<'ctx> From<ContiguousNDArrayType<'ctx>> for PointerType<'ctx> {
fn from(value: ContiguousNDArrayType<'ctx>) -> Self {
value.as_base_type()
}
}

View File

@ -0,0 +1,215 @@
use inkwell::{
context::{AsContextRef, Context},
types::{AnyTypeEnum, BasicType, BasicTypeEnum, IntType, PointerType},
values::{IntValue, PointerValue},
AddressSpace,
};
use itertools::Itertools;
use nac3core_derive::StructFields;
use crate::codegen::{
types::{
structure::{check_struct_type_matches_fields, StructField, StructFields},
ProxyType,
},
values::{
ndarray::{NDIndexValue, RustNDIndex},
ArrayLikeIndexer, ArraySliceValue, ProxyValue,
},
CodeGenContext, CodeGenerator,
};
#[derive(Debug, PartialEq, Eq, Clone, Copy)]
pub struct NDIndexType<'ctx> {
ty: PointerType<'ctx>,
llvm_usize: IntType<'ctx>,
}
#[derive(PartialEq, Eq, Clone, Copy, StructFields)]
pub struct NDIndexStructFields<'ctx> {
#[value_type(i8_type())]
pub type_: StructField<'ctx, IntValue<'ctx>>,
#[value_type(i8_type().ptr_type(AddressSpace::default()))]
pub data: StructField<'ctx, PointerValue<'ctx>>,
}
impl<'ctx> NDIndexType<'ctx> {
/// Checks whether `llvm_ty` represents a `ndindex` type, returning [Err] if it does not.
pub fn is_representable(
llvm_ty: PointerType<'ctx>,
llvm_usize: IntType<'ctx>,
) -> Result<(), String> {
let ctx = llvm_ty.get_context();
let llvm_ty = llvm_ty.get_element_type();
let AnyTypeEnum::StructType(llvm_ty) = llvm_ty else {
return Err(format!(
"Expected struct type for `ContiguousNDArray` type, got {llvm_ty}"
));
};
let fields = NDIndexStructFields::new(ctx, llvm_usize);
check_struct_type_matches_fields(fields, llvm_ty, "NDIndex", &[])
}
#[must_use]
fn fields(
ctx: impl AsContextRef<'ctx>,
llvm_usize: IntType<'ctx>,
) -> NDIndexStructFields<'ctx> {
NDIndexStructFields::new(ctx, llvm_usize)
}
#[must_use]
pub fn get_fields(&self) -> NDIndexStructFields<'ctx> {
Self::fields(self.ty.get_context(), self.llvm_usize)
}
#[must_use]
fn llvm_type(ctx: &'ctx Context, llvm_usize: IntType<'ctx>) -> PointerType<'ctx> {
let field_tys =
Self::fields(ctx, llvm_usize).into_iter().map(|field| field.1).collect_vec();
ctx.struct_type(&field_tys, false).ptr_type(AddressSpace::default())
}
#[must_use]
pub fn new<G: CodeGenerator + ?Sized>(generator: &G, ctx: &'ctx Context) -> Self {
let llvm_usize = generator.get_size_type(ctx);
let llvm_ndindex = Self::llvm_type(ctx, llvm_usize);
Self { ty: llvm_ndindex, llvm_usize }
}
#[must_use]
pub fn from_type(ptr_ty: PointerType<'ctx>, llvm_usize: IntType<'ctx>) -> Self {
debug_assert!(Self::is_representable(ptr_ty, llvm_usize).is_ok());
Self { ty: ptr_ty, llvm_usize }
}
#[must_use]
pub fn alloca<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
name: Option<&'ctx str>,
) -> <Self as ProxyType<'ctx>>::Value {
<Self as ProxyType<'ctx>>::Value::from_pointer_value(
self.raw_alloca(generator, ctx, name),
self.llvm_usize,
name,
)
}
/// Serialize a list of [`RustNDIndex`] as a newly allocated LLVM array of [`NDIndexValue`].
#[must_use]
pub fn construct_ndindices<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
in_ndindices: &[RustNDIndex<'ctx>],
) -> ArraySliceValue<'ctx> {
// Allocate the LLVM ndindices.
let num_ndindices = self.llvm_usize.const_int(in_ndindices.len() as u64, false);
let ndindices = self.array_alloca(generator, ctx, num_ndindices, None);
// Initialize all of them.
for (i, in_ndindex) in in_ndindices.iter().enumerate() {
let pndindex = unsafe {
ndindices.ptr_offset_unchecked(
ctx,
generator,
&ctx.ctx.i64_type().const_int(u64::try_from(i).unwrap(), false),
None,
)
};
in_ndindex.write_to_ndindex(
generator,
ctx,
NDIndexValue::from_pointer_value(pndindex, self.llvm_usize, None),
);
}
ndindices
}
#[must_use]
pub fn map_value(
&self,
value: <<Self as ProxyType<'ctx>>::Value as ProxyValue<'ctx>>::Base,
name: Option<&'ctx str>,
) -> <Self as ProxyType<'ctx>>::Value {
<Self as ProxyType<'ctx>>::Value::from_pointer_value(value, self.llvm_usize, name)
}
}
impl<'ctx> ProxyType<'ctx> for NDIndexType<'ctx> {
type Base = PointerType<'ctx>;
type Value = NDIndexValue<'ctx>;
fn is_type<G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &'ctx Context,
llvm_ty: impl BasicType<'ctx>,
) -> Result<(), String> {
if let BasicTypeEnum::PointerType(ty) = llvm_ty.as_basic_type_enum() {
<Self as ProxyType<'ctx>>::is_representable(generator, ctx, ty)
} else {
Err(format!("Expected pointer type, got {llvm_ty:?}"))
}
}
fn is_representable<G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &'ctx Context,
llvm_ty: Self::Base,
) -> Result<(), String> {
Self::is_representable(llvm_ty, generator.get_size_type(ctx))
}
fn raw_alloca<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
name: Option<&'ctx str>,
) -> <Self::Value as ProxyValue<'ctx>>::Base {
generator
.gen_var_alloc(
ctx,
self.as_base_type().get_element_type().into_struct_type().into(),
name,
)
.unwrap()
}
fn array_alloca<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
size: IntValue<'ctx>,
name: Option<&'ctx str>,
) -> ArraySliceValue<'ctx> {
generator
.gen_array_var_alloc(
ctx,
self.as_base_type().get_element_type().into_struct_type().into(),
size,
name,
)
.unwrap()
}
fn as_base_type(&self) -> Self::Base {
self.ty
}
}
impl<'ctx> From<NDIndexType<'ctx>> for PointerType<'ctx> {
fn from(value: NDIndexType<'ctx>) -> Self {
value.as_base_type()
}
}

View File

@ -0,0 +1,469 @@
use inkwell::{
context::{AsContextRef, Context},
types::{AnyTypeEnum, BasicType, BasicTypeEnum, IntType, PointerType},
values::{BasicValue, IntValue, PointerValue},
AddressSpace,
};
use itertools::Itertools;
use nac3core_derive::StructFields;
use super::{
structure::{check_struct_type_matches_fields, StructField, StructFields},
ProxyType,
};
use crate::{
codegen::{
values::{ndarray::NDArrayValue, ArraySliceValue, ProxyValue, TypedArrayLikeMutator},
{CodeGenContext, CodeGenerator},
},
toplevel::{helper::extract_ndims, numpy::unpack_ndarray_var_tys},
typecheck::typedef::Type,
};
pub use contiguous::*;
pub use indexing::*;
pub use nditer::*;
mod contiguous;
mod indexing;
mod nditer;
/// Proxy type for a `ndarray` type in LLVM.
#[derive(Debug, PartialEq, Eq, Clone, Copy)]
pub struct NDArrayType<'ctx> {
ty: PointerType<'ctx>,
dtype: BasicTypeEnum<'ctx>,
ndims: Option<u64>,
llvm_usize: IntType<'ctx>,
}
#[derive(PartialEq, Eq, Clone, Copy, StructFields)]
pub struct NDArrayStructFields<'ctx> {
/// The size of each `NDArray` element in bytes.
#[value_type(usize)]
pub itemsize: StructField<'ctx, IntValue<'ctx>>,
/// Number of dimensions in the array.
#[value_type(usize)]
pub ndims: StructField<'ctx, IntValue<'ctx>>,
/// Pointer to an array containing the shape of the `NDArray`.
#[value_type(usize.ptr_type(AddressSpace::default()))]
pub shape: StructField<'ctx, PointerValue<'ctx>>,
/// Pointer to an array indicating the number of bytes between each element at a dimension
#[value_type(usize.ptr_type(AddressSpace::default()))]
pub strides: StructField<'ctx, PointerValue<'ctx>>,
/// Pointer to an array containing the array data
#[value_type(i8_type().ptr_type(AddressSpace::default()))]
pub data: StructField<'ctx, PointerValue<'ctx>>,
}
impl<'ctx> NDArrayType<'ctx> {
/// Checks whether `llvm_ty` represents a `ndarray` type, returning [Err] if it does not.
pub fn is_representable(
llvm_ty: PointerType<'ctx>,
llvm_usize: IntType<'ctx>,
) -> Result<(), String> {
let ctx = llvm_ty.get_context();
let llvm_ndarray_ty = llvm_ty.get_element_type();
let AnyTypeEnum::StructType(llvm_ndarray_ty) = llvm_ndarray_ty else {
return Err(format!("Expected struct type for `NDArray` type, got {llvm_ndarray_ty}"));
};
check_struct_type_matches_fields(
Self::fields(ctx, llvm_usize),
llvm_ndarray_ty,
"NDArray",
&[],
)
}
/// Returns an instance of [`StructFields`] containing all field accessors for this type.
#[must_use]
fn fields(
ctx: impl AsContextRef<'ctx>,
llvm_usize: IntType<'ctx>,
) -> NDArrayStructFields<'ctx> {
NDArrayStructFields::new(ctx, llvm_usize)
}
/// See [`NDArrayType::fields`].
// TODO: Move this into e.g. StructProxyType
#[must_use]
pub fn get_fields(&self, ctx: impl AsContextRef<'ctx>) -> NDArrayStructFields<'ctx> {
Self::fields(ctx, self.llvm_usize)
}
/// Creates an LLVM type corresponding to the expected structure of an `NDArray`.
#[must_use]
fn llvm_type(ctx: &'ctx Context, llvm_usize: IntType<'ctx>) -> PointerType<'ctx> {
let field_tys =
Self::fields(ctx, llvm_usize).into_iter().map(|field| field.1).collect_vec();
ctx.struct_type(&field_tys, false).ptr_type(AddressSpace::default())
}
/// Creates an instance of [`NDArrayType`].
#[must_use]
pub fn new<G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &'ctx Context,
dtype: BasicTypeEnum<'ctx>,
ndims: Option<u64>,
) -> Self {
let llvm_usize = generator.get_size_type(ctx);
let llvm_ndarray = Self::llvm_type(ctx, llvm_usize);
NDArrayType { ty: llvm_ndarray, dtype, ndims, llvm_usize }
}
/// Creates an instance of [`NDArrayType`] with `ndims` of 0.
#[must_use]
pub fn new_unsized<G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &'ctx Context,
dtype: BasicTypeEnum<'ctx>,
) -> Self {
let llvm_usize = generator.get_size_type(ctx);
let llvm_ndarray = Self::llvm_type(ctx, llvm_usize);
NDArrayType { ty: llvm_ndarray, dtype, ndims: Some(0), llvm_usize }
}
/// Creates an [`NDArrayType`] from a [unifier type][Type].
#[must_use]
pub fn from_unifier_type<G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &mut CodeGenContext<'ctx, '_>,
ty: Type,
) -> Self {
let (dtype, ndims) = unpack_ndarray_var_tys(&mut ctx.unifier, ty);
let llvm_dtype = ctx.get_llvm_type(generator, dtype);
let llvm_usize = generator.get_size_type(ctx.ctx);
let ndims = extract_ndims(&ctx.unifier, ndims);
NDArrayType {
ty: Self::llvm_type(ctx.ctx, llvm_usize),
dtype: llvm_dtype,
ndims: Some(ndims),
llvm_usize,
}
}
/// Creates an [`NDArrayType`] from a [`PointerType`] representing an `NDArray`.
#[must_use]
pub fn from_type(
ptr_ty: PointerType<'ctx>,
dtype: BasicTypeEnum<'ctx>,
ndims: Option<u64>,
llvm_usize: IntType<'ctx>,
) -> Self {
debug_assert!(Self::is_representable(ptr_ty, llvm_usize).is_ok());
NDArrayType { ty: ptr_ty, dtype, ndims, llvm_usize }
}
/// Returns the type of the `size` field of this `ndarray` type.
#[must_use]
pub fn size_type(&self) -> IntType<'ctx> {
self.llvm_usize
}
/// Returns the element type of this `ndarray` type.
#[must_use]
pub fn element_type(&self) -> BasicTypeEnum<'ctx> {
self.dtype
}
/// Returns the number of dimensions of this `ndarray` type.
#[must_use]
pub fn ndims(&self) -> Option<u64> {
self.ndims
}
/// Allocates an instance of [`NDArrayValue`] as if by calling `alloca` on the base type.
#[must_use]
pub fn alloca<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
name: Option<&'ctx str>,
) -> <Self as ProxyType<'ctx>>::Value {
<Self as ProxyType<'ctx>>::Value::from_pointer_value(
self.raw_alloca(generator, ctx, name),
self.dtype,
self.ndims,
self.llvm_usize,
name,
)
}
/// Allocates an [`NDArrayValue`] on the stack and initializes all fields as follows:
///
/// - `data`: uninitialized.
/// - `itemsize`: set to the size of `self.dtype`.
/// - `ndims`: set to the value of `ndims`.
/// - `shape`: allocated on the stack with an array of length `ndims` with uninitialized values.
/// - `strides`: allocated on the stack with an array of length `ndims` with uninitialized
/// values.
#[must_use]
fn construct_impl<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
ndims: IntValue<'ctx>,
name: Option<&'ctx str>,
) -> <Self as ProxyType<'ctx>>::Value {
let ndarray = self.alloca(generator, ctx, name);
let itemsize = ctx
.builder
.build_int_truncate_or_bit_cast(self.dtype.size_of().unwrap(), self.llvm_usize, "")
.unwrap();
ndarray.store_itemsize(ctx, generator, itemsize);
ndarray.store_ndims(ctx, generator, ndims);
ndarray.create_shape(ctx, self.llvm_usize, ndims);
ndarray.create_strides(ctx, self.llvm_usize, ndims);
ndarray
}
/// Allocate an [`NDArrayValue`] on the stack using `dtype` and `ndims` of this [`NDArrayType`]
/// instance.
///
/// The returned ndarray's content will be:
/// - `data`: uninitialized.
/// - `itemsize`: set to the size of `dtype`.
/// - `ndims`: set to the value of `self.ndims`.
/// - `shape`: allocated on the stack with an array of length `ndims` with uninitialized values.
/// - `strides`: allocated on the stack with an array of length `ndims` with uninitialized
/// values.
#[must_use]
pub fn construct_uninitialized<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
name: Option<&'ctx str>,
) -> <Self as ProxyType<'ctx>>::Value {
assert!(self.ndims.is_some(), "NDArrayType::construct can only be called on an instance with compile-time known ndims (self.ndims = Some(ndims))");
let Some(ndims) = self.ndims.map(|ndims| self.llvm_usize.const_int(ndims, false)) else {
unreachable!()
};
self.construct_impl(generator, ctx, ndims, name)
}
/// Allocate an [`NDArrayValue`] on the stack given its `ndims` and `dtype`.
///
/// `shape` and `strides` will be automatically allocated onto the stack.
///
/// The returned ndarray's content will be:
/// - `data`: uninitialized.
/// - `itemsize`: set to the size of `dtype`.
/// - `ndims`: set to the value of `ndims`.
/// - `shape`: allocated with an array of length `ndims` with uninitialized values.
/// - `strides`: allocated with an array of length `ndims` with uninitialized values.
#[deprecated = "Prefer construct_uninitialized or construct_*_shape."]
#[must_use]
pub fn construct_dyn_ndims<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
ndims: IntValue<'ctx>,
name: Option<&'ctx str>,
) -> <Self as ProxyType<'ctx>>::Value {
assert!(self.ndims.is_none(), "NDArrayType::construct_dyn_ndims can only be called on an instance with compile-time unknown ndims (self.ndims = None)");
self.construct_impl(generator, ctx, ndims, name)
}
/// Convenience function. Allocate an [`NDArrayValue`] with a statically known shape.
///
/// The returned [`NDArrayValue`]'s `data` and `strides` are uninitialized.
#[must_use]
pub fn construct_const_shape<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
shape: &[u64],
name: Option<&'ctx str>,
) -> <Self as ProxyType<'ctx>>::Value {
assert!(self.ndims.is_none_or(|ndims| shape.len() as u64 == ndims));
let ndarray = Self::new(generator, ctx.ctx, self.dtype, Some(shape.len() as u64))
.construct_uninitialized(generator, ctx, name);
let llvm_usize = generator.get_size_type(ctx.ctx);
// Write shape
let ndarray_shape = ndarray.shape();
for (i, dim) in shape.iter().enumerate() {
let dim = llvm_usize.const_int(*dim, false);
unsafe {
ndarray_shape.set_typed_unchecked(
ctx,
generator,
&llvm_usize.const_int(i as u64, false),
dim,
);
}
}
ndarray
}
/// Convenience function. Allocate an [`NDArrayValue`] with a dynamically known shape.
///
/// The returned [`NDArrayValue`]'s `data` and `strides` are uninitialized.
#[must_use]
pub fn construct_dyn_shape<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
shape: &[IntValue<'ctx>],
name: Option<&'ctx str>,
) -> <Self as ProxyType<'ctx>>::Value {
assert!(self.ndims.is_none_or(|ndims| shape.len() as u64 == ndims));
let ndarray = Self::new(generator, ctx.ctx, self.dtype, Some(shape.len() as u64))
.construct_uninitialized(generator, ctx, name);
let llvm_usize = generator.get_size_type(ctx.ctx);
// Write shape
let ndarray_shape = ndarray.shape();
for (i, dim) in shape.iter().enumerate() {
assert_eq!(
dim.get_type(),
llvm_usize,
"Expected {} but got {}",
llvm_usize.print_to_string(),
dim.get_type().print_to_string()
);
unsafe {
ndarray_shape.set_typed_unchecked(
ctx,
generator,
&llvm_usize.const_int(i as u64, false),
*dim,
);
}
}
ndarray
}
/// Create an unsized ndarray to contain `value`.
#[must_use]
pub fn construct_unsized<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
value: &impl BasicValue<'ctx>,
name: Option<&'ctx str>,
) -> NDArrayValue<'ctx> {
let value = value.as_basic_value_enum();
assert_eq!(value.get_type(), self.dtype);
assert!(self.ndims.is_none_or(|ndims| ndims == 0));
// We have to put the value on the stack to get a data pointer.
let data = ctx.builder.build_alloca(value.get_type(), "construct_unsized").unwrap();
ctx.builder.build_store(data, value).unwrap();
let data = ctx
.builder
.build_pointer_cast(data, ctx.ctx.i8_type().ptr_type(AddressSpace::default()), "")
.unwrap();
let ndarray = Self::new_unsized(generator, ctx.ctx, value.get_type())
.construct_uninitialized(generator, ctx, name);
ctx.builder.build_store(ndarray.ptr_to_data(ctx), data).unwrap();
ndarray
}
/// Converts an existing value into a [`NDArrayValue`].
#[must_use]
pub fn map_value(
&self,
value: <<Self as ProxyType<'ctx>>::Value as ProxyValue<'ctx>>::Base,
name: Option<&'ctx str>,
) -> <Self as ProxyType<'ctx>>::Value {
<Self as ProxyType<'ctx>>::Value::from_pointer_value(
value,
self.dtype,
self.ndims,
self.llvm_usize,
name,
)
}
}
impl<'ctx> ProxyType<'ctx> for NDArrayType<'ctx> {
type Base = PointerType<'ctx>;
type Value = NDArrayValue<'ctx>;
fn is_type<G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &'ctx Context,
llvm_ty: impl BasicType<'ctx>,
) -> Result<(), String> {
if let BasicTypeEnum::PointerType(ty) = llvm_ty.as_basic_type_enum() {
<Self as ProxyType<'ctx>>::is_representable(generator, ctx, ty)
} else {
Err(format!("Expected pointer type, got {llvm_ty:?}"))
}
}
fn is_representable<G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &'ctx Context,
llvm_ty: Self::Base,
) -> Result<(), String> {
Self::is_representable(llvm_ty, generator.get_size_type(ctx))
}
fn raw_alloca<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
name: Option<&'ctx str>,
) -> <Self::Value as ProxyValue<'ctx>>::Base {
generator
.gen_var_alloc(
ctx,
self.as_base_type().get_element_type().into_struct_type().into(),
name,
)
.unwrap()
}
fn array_alloca<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
size: IntValue<'ctx>,
name: Option<&'ctx str>,
) -> ArraySliceValue<'ctx> {
generator
.gen_array_var_alloc(
ctx,
self.as_base_type().get_element_type().into_struct_type().into(),
size,
name,
)
.unwrap()
}
fn as_base_type(&self) -> Self::Base {
self.ty
}
}
impl<'ctx> From<NDArrayType<'ctx>> for PointerType<'ctx> {
fn from(value: NDArrayType<'ctx>) -> Self {
value.as_base_type()
}
}

View File

@ -0,0 +1,241 @@
use inkwell::{
context::{AsContextRef, Context},
types::{AnyTypeEnum, BasicType, BasicTypeEnum, IntType, PointerType},
values::{IntValue, PointerValue},
AddressSpace,
};
use itertools::Itertools;
use nac3core_derive::StructFields;
use super::ProxyType;
use crate::codegen::{
irrt,
types::structure::{check_struct_type_matches_fields, StructField, StructFields},
values::{
ndarray::{NDArrayValue, NDIterValue},
ArraySliceValue, ProxyValue,
},
CodeGenContext, CodeGenerator,
};
#[derive(Debug, PartialEq, Eq, Clone, Copy)]
pub struct NDIterType<'ctx> {
ty: PointerType<'ctx>,
llvm_usize: IntType<'ctx>,
}
#[derive(PartialEq, Eq, Clone, Copy, StructFields)]
pub struct NDIterStructFields<'ctx> {
#[value_type(usize)]
pub ndims: StructField<'ctx, IntValue<'ctx>>,
#[value_type(usize.ptr_type(AddressSpace::default()))]
pub shape: StructField<'ctx, PointerValue<'ctx>>,
#[value_type(usize.ptr_type(AddressSpace::default()))]
pub strides: StructField<'ctx, PointerValue<'ctx>>,
#[value_type(usize.ptr_type(AddressSpace::default()))]
pub indices: StructField<'ctx, PointerValue<'ctx>>,
#[value_type(usize)]
pub nth: StructField<'ctx, IntValue<'ctx>>,
#[value_type(i8_type().ptr_type(AddressSpace::default()))]
pub element: StructField<'ctx, PointerValue<'ctx>>,
#[value_type(usize)]
pub size: StructField<'ctx, IntValue<'ctx>>,
}
impl<'ctx> NDIterType<'ctx> {
/// Checks whether `llvm_ty` represents a `nditer` type, returning [Err] if it does not.
pub fn is_representable(
llvm_ty: PointerType<'ctx>,
llvm_usize: IntType<'ctx>,
) -> Result<(), String> {
let ctx = llvm_ty.get_context();
let llvm_ty = llvm_ty.get_element_type();
let AnyTypeEnum::StructType(llvm_ndarray_ty) = llvm_ty else {
return Err(format!("Expected struct type for `NDIter` type, got {llvm_ty}"));
};
check_struct_type_matches_fields(
Self::fields(ctx, llvm_usize),
llvm_ndarray_ty,
"NDIter",
&[],
)
}
/// Returns an instance of [`StructFields`] containing all field accessors for this type.
#[must_use]
fn fields(ctx: impl AsContextRef<'ctx>, llvm_usize: IntType<'ctx>) -> NDIterStructFields<'ctx> {
NDIterStructFields::new(ctx, llvm_usize)
}
/// See [`NDIterType::fields`].
// TODO: Move this into e.g. StructProxyType
#[must_use]
pub fn get_fields(&self, ctx: impl AsContextRef<'ctx>) -> NDIterStructFields<'ctx> {
Self::fields(ctx, self.llvm_usize)
}
/// Creates an LLVM type corresponding to the expected structure of an `NDIter`.
#[must_use]
fn llvm_type(ctx: &'ctx Context, llvm_usize: IntType<'ctx>) -> PointerType<'ctx> {
let field_tys =
Self::fields(ctx, llvm_usize).into_iter().map(|field| field.1).collect_vec();
ctx.struct_type(&field_tys, false).ptr_type(AddressSpace::default())
}
/// Creates an instance of [`NDIter`].
#[must_use]
pub fn new<G: CodeGenerator + ?Sized>(generator: &G, ctx: &'ctx Context) -> Self {
let llvm_usize = generator.get_size_type(ctx);
let llvm_nditer = Self::llvm_type(ctx, llvm_usize);
Self { ty: llvm_nditer, llvm_usize }
}
/// Creates an [`NDIterType`] from a [`PointerType`] representing an `NDIter`.
#[must_use]
pub fn from_type(ptr_ty: PointerType<'ctx>, llvm_usize: IntType<'ctx>) -> Self {
debug_assert!(Self::is_representable(ptr_ty, llvm_usize).is_ok());
Self { ty: ptr_ty, llvm_usize }
}
/// Returns the type of the `size` field of this `nditer` type.
#[must_use]
pub fn size_type(&self) -> IntType<'ctx> {
self.llvm_usize
}
#[must_use]
pub fn alloca<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
parent: NDArrayValue<'ctx>,
indices: ArraySliceValue<'ctx>,
name: Option<&'ctx str>,
) -> <Self as ProxyType<'ctx>>::Value {
<Self as ProxyType<'ctx>>::Value::from_pointer_value(
self.raw_alloca(generator, ctx, name),
parent,
indices,
self.llvm_usize,
name,
)
}
/// Allocate an [`NDIter`] that iterates through the given `ndarray`.
#[must_use]
pub fn construct<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
ndarray: NDArrayValue<'ctx>,
) -> <Self as ProxyType<'ctx>>::Value {
let nditer = self.raw_alloca(generator, ctx, None);
let ndims = ndarray.load_ndims(ctx);
// The caller has the responsibility to allocate 'indices' for `NDIter`.
let indices =
generator.gen_array_var_alloc(ctx, self.llvm_usize.into(), ndims, None).unwrap();
let nditer = <Self as ProxyType<'ctx>>::Value::from_pointer_value(
nditer,
ndarray,
indices,
self.llvm_usize,
None,
);
irrt::ndarray::call_nac3_nditer_initialize(generator, ctx, nditer, ndarray, indices);
nditer
}
#[must_use]
pub fn map_value(
&self,
value: <<Self as ProxyType<'ctx>>::Value as ProxyValue<'ctx>>::Base,
parent: NDArrayValue<'ctx>,
indices: ArraySliceValue<'ctx>,
name: Option<&'ctx str>,
) -> <Self as ProxyType<'ctx>>::Value {
<Self as ProxyType<'ctx>>::Value::from_pointer_value(
value,
parent,
indices,
self.llvm_usize,
name,
)
}
}
impl<'ctx> ProxyType<'ctx> for NDIterType<'ctx> {
type Base = PointerType<'ctx>;
type Value = NDIterValue<'ctx>;
fn is_type<G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &'ctx Context,
llvm_ty: impl BasicType<'ctx>,
) -> Result<(), String> {
if let BasicTypeEnum::PointerType(ty) = llvm_ty.as_basic_type_enum() {
<Self as ProxyType<'ctx>>::is_representable(generator, ctx, ty)
} else {
Err(format!("Expected pointer type, got {llvm_ty:?}"))
}
}
fn is_representable<G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &'ctx Context,
llvm_ty: Self::Base,
) -> Result<(), String> {
Self::is_representable(llvm_ty, generator.get_size_type(ctx))
}
fn raw_alloca<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
name: Option<&'ctx str>,
) -> <Self::Value as ProxyValue<'ctx>>::Base {
generator
.gen_var_alloc(
ctx,
self.as_base_type().get_element_type().into_struct_type().into(),
name,
)
.unwrap()
}
fn array_alloca<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
size: IntValue<'ctx>,
name: Option<&'ctx str>,
) -> ArraySliceValue<'ctx> {
generator
.gen_array_var_alloc(
ctx,
self.as_base_type().get_element_type().into_struct_type().into(),
size,
name,
)
.unwrap()
}
fn as_base_type(&self) -> Self::Base {
self.ty
}
}
impl<'ctx> From<NDIterType<'ctx>> for PointerType<'ctx> {
fn from(value: NDIterType<'ctx>) -> Self {
value.as_base_type()
}
}

View File

@ -0,0 +1,170 @@
use inkwell::{
context::Context,
types::{AnyTypeEnum, BasicType, BasicTypeEnum, IntType, PointerType},
values::IntValue,
AddressSpace,
};
use super::ProxyType;
use crate::codegen::{
values::{ArraySliceValue, ProxyValue, RangeValue},
{CodeGenContext, CodeGenerator},
};
/// Proxy type for a `range` type in LLVM.
#[derive(Debug, PartialEq, Eq, Clone, Copy)]
pub struct RangeType<'ctx> {
ty: PointerType<'ctx>,
}
impl<'ctx> RangeType<'ctx> {
/// Checks whether `llvm_ty` represents a `range` type, returning [Err] if it does not.
pub fn is_representable(llvm_ty: PointerType<'ctx>) -> Result<(), String> {
let llvm_range_ty = llvm_ty.get_element_type();
let AnyTypeEnum::ArrayType(llvm_range_ty) = llvm_range_ty else {
return Err(format!("Expected array type for `range` type, got {llvm_range_ty}"));
};
if llvm_range_ty.len() != 3 {
return Err(format!(
"Expected 3 elements for `range` type, got {}",
llvm_range_ty.len()
));
}
let llvm_range_elem_ty = llvm_range_ty.get_element_type();
let Ok(llvm_range_elem_ty) = IntType::try_from(llvm_range_elem_ty) else {
return Err(format!(
"Expected int type for `range` element type, got {llvm_range_elem_ty}"
));
};
if llvm_range_elem_ty.get_bit_width() != 32 {
return Err(format!(
"Expected 32-bit int type for `range` element type, got {}",
llvm_range_elem_ty.get_bit_width()
));
}
Ok(())
}
/// Creates an LLVM type corresponding to the expected structure of a `Range`.
#[must_use]
fn llvm_type(ctx: &'ctx Context) -> PointerType<'ctx> {
// typedef int32_t Range[3];
let llvm_i32 = ctx.i32_type();
llvm_i32.array_type(3).ptr_type(AddressSpace::default())
}
/// Creates an instance of [`RangeType`].
#[must_use]
pub fn new(ctx: &'ctx Context) -> Self {
let llvm_range = Self::llvm_type(ctx);
RangeType::from_type(llvm_range)
}
/// Creates an [`RangeType`] from a [`PointerType`].
#[must_use]
pub fn from_type(ptr_ty: PointerType<'ctx>) -> Self {
debug_assert!(Self::is_representable(ptr_ty).is_ok());
RangeType { ty: ptr_ty }
}
/// Returns the type of all fields of this `range` type.
#[must_use]
pub fn value_type(&self) -> IntType<'ctx> {
self.as_base_type().get_element_type().into_array_type().get_element_type().into_int_type()
}
/// Allocates an instance of [`RangeValue`] as if by calling `alloca` on the base type.
#[must_use]
pub fn alloca<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
name: Option<&'ctx str>,
) -> <Self as ProxyType<'ctx>>::Value {
<Self as ProxyType<'ctx>>::Value::from_pointer_value(
self.raw_alloca(generator, ctx, name),
name,
)
}
/// Converts an existing value into a [`RangeValue`].
#[must_use]
pub fn map_value(
&self,
value: <<Self as ProxyType<'ctx>>::Value as ProxyValue<'ctx>>::Base,
name: Option<&'ctx str>,
) -> <Self as ProxyType<'ctx>>::Value {
<Self as ProxyType<'ctx>>::Value::from_pointer_value(value, name)
}
}
impl<'ctx> ProxyType<'ctx> for RangeType<'ctx> {
type Base = PointerType<'ctx>;
type Value = RangeValue<'ctx>;
fn is_type<G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &'ctx Context,
llvm_ty: impl BasicType<'ctx>,
) -> Result<(), String> {
if let BasicTypeEnum::PointerType(ty) = llvm_ty.as_basic_type_enum() {
<Self as ProxyType<'ctx>>::is_representable(generator, ctx, ty)
} else {
Err(format!("Expected pointer type, got {llvm_ty:?}"))
}
}
fn is_representable<G: CodeGenerator + ?Sized>(
_: &G,
_: &'ctx Context,
llvm_ty: Self::Base,
) -> Result<(), String> {
Self::is_representable(llvm_ty)
}
fn raw_alloca<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
name: Option<&'ctx str>,
) -> <Self::Value as ProxyValue<'ctx>>::Base {
generator
.gen_var_alloc(
ctx,
self.as_base_type().get_element_type().into_struct_type().into(),
name,
)
.unwrap()
}
fn array_alloca<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
size: IntValue<'ctx>,
name: Option<&'ctx str>,
) -> ArraySliceValue<'ctx> {
generator
.gen_array_var_alloc(
ctx,
self.as_base_type().get_element_type().into_struct_type().into(),
size,
name,
)
.unwrap()
}
fn as_base_type(&self) -> Self::Base {
self.ty
}
}
impl<'ctx> From<RangeType<'ctx>> for PointerType<'ctx> {
fn from(value: RangeType<'ctx>) -> Self {
value.as_base_type()
}
}

View File

@ -0,0 +1,255 @@
use std::marker::PhantomData;
use inkwell::{
context::AsContextRef,
types::{BasicTypeEnum, IntType, StructType},
values::{BasicValue, BasicValueEnum, IntValue, PointerValue, StructValue},
};
use crate::codegen::CodeGenContext;
/// Trait indicating that the structure is a field-wise representation of an LLVM structure.
///
/// # Usage
///
/// For example, for a simple C-slice LLVM structure:
///
/// ```ignore
/// struct CSliceFields<'ctx> {
/// ptr: StructField<'ctx, PointerValue<'ctx>>,
/// len: StructField<'ctx, IntValue<'ctx>>
/// }
/// ```
pub trait StructFields<'ctx>: Eq + Copy {
/// Creates an instance of [`StructFields`] using the given `ctx` and `size_t` types.
fn new(ctx: impl AsContextRef<'ctx>, llvm_usize: IntType<'ctx>) -> Self;
/// Returns a [`Vec`] that contains the fields of the structure in the order as they appear in
/// the type definition.
#[must_use]
fn to_vec(&self) -> Vec<(&'static str, BasicTypeEnum<'ctx>)>;
/// Returns a [`Iterator`] that contains the fields of the structure in the order as they appear
/// in the type definition.
#[must_use]
fn iter(&self) -> impl Iterator<Item = (&'static str, BasicTypeEnum<'ctx>)> {
self.to_vec().into_iter()
}
/// Returns a [`Vec`] that contains the fields of the structure in the order as they appear in
/// the type definition.
#[must_use]
fn into_vec(self) -> Vec<(&'static str, BasicTypeEnum<'ctx>)>
where
Self: Sized,
{
self.to_vec()
}
/// Returns a [`Iterator`] that contains the fields of the structure in the order as they appear
/// in the type definition.
#[must_use]
fn into_iter(self) -> impl Iterator<Item = (&'static str, BasicTypeEnum<'ctx>)>
where
Self: Sized,
{
self.into_vec().into_iter()
}
}
/// A single field of an LLVM structure.
#[derive(Debug, PartialEq, Eq, Clone, Copy)]
pub struct StructField<'ctx, Value>
where
Value: BasicValue<'ctx> + TryFrom<BasicValueEnum<'ctx>, Error = ()>,
{
/// The index of this field within the structure.
index: u32,
/// The name of this field.
name: &'static str,
/// The type of this field.
ty: BasicTypeEnum<'ctx>,
/// Instance of [`PhantomData`] containing [`Value`], used to implement automatic downcasts.
_value_ty: PhantomData<Value>,
}
impl<'ctx, Value> StructField<'ctx, Value>
where
Value: BasicValue<'ctx> + TryFrom<BasicValueEnum<'ctx>, Error = ()>,
{
/// Creates an instance of [`StructField`].
///
/// * `idx_counter` - The instance of [`FieldIndexCounter`] used to track the current field
/// index.
/// * `name` - Name of the field.
/// * `ty` - The type of this field.
pub fn create(
idx_counter: &mut FieldIndexCounter,
name: &'static str,
ty: impl Into<BasicTypeEnum<'ctx>>,
) -> Self {
StructField { index: idx_counter.increment(), name, ty: ty.into(), _value_ty: PhantomData }
}
/// Creates an instance of [`StructField`] with a given index.
///
/// * `index` - The index of this field within its enclosing structure.
/// * `name` - Name of the field.
/// * `ty` - The type of this field.
pub fn create_at(index: u32, name: &'static str, ty: impl Into<BasicTypeEnum<'ctx>>) -> Self {
StructField { index, name, ty: ty.into(), _value_ty: PhantomData }
}
/// Returns the name of this field.
#[must_use]
pub fn name(&self) -> &'static str {
self.name
}
/// Creates a pointer to this field in an arbitrary structure by performing a `getelementptr i32
/// {idx...}, i32 {self.index}`.
pub fn ptr_by_array_gep(
&self,
ctx: &CodeGenContext<'ctx, '_>,
pobj: PointerValue<'ctx>,
idx: &[IntValue<'ctx>],
) -> PointerValue<'ctx> {
unsafe {
ctx.builder.build_in_bounds_gep(
pobj,
&[idx, &[ctx.ctx.i32_type().const_int(u64::from(self.index), false)]].concat(),
"",
)
}
.unwrap()
}
/// Creates a pointer to this field in an arbitrary structure by performing the equivalent of
/// `getelementptr i32 0, i32 {self.index}`.
pub fn ptr_by_gep(
&self,
ctx: &CodeGenContext<'ctx, '_>,
pobj: PointerValue<'ctx>,
obj_name: Option<&'ctx str>,
) -> PointerValue<'ctx> {
ctx.builder
.build_struct_gep(
pobj,
self.index,
&obj_name.map(|name| format!("{name}.{}.addr", self.name)).unwrap_or_default(),
)
.unwrap()
}
/// Gets the value of this field for a given `obj`.
#[must_use]
pub fn get_from_value(&self, obj: StructValue<'ctx>) -> Value {
obj.get_field_at_index(self.index).and_then(|value| Value::try_from(value).ok()).unwrap()
}
/// Sets the value of this field for a given `obj`.
pub fn set_for_value(&self, obj: StructValue<'ctx>, value: Value) {
obj.set_field_at_index(self.index, value);
}
/// Gets the value of this field for a pointer-to-structure.
pub fn get(
&self,
ctx: &CodeGenContext<'ctx, '_>,
pobj: PointerValue<'ctx>,
obj_name: Option<&'ctx str>,
) -> Value {
ctx.builder
.build_load(
self.ptr_by_gep(ctx, pobj, obj_name),
&obj_name.map(|name| format!("{name}.{}", self.name)).unwrap_or_default(),
)
.map_err(|_| ())
.and_then(|value| Value::try_from(value))
.unwrap()
}
/// Sets the value of this field for a pointer-to-structure.
pub fn set(
&self,
ctx: &CodeGenContext<'ctx, '_>,
pobj: PointerValue<'ctx>,
value: Value,
obj_name: Option<&'ctx str>,
) {
ctx.builder.build_store(self.ptr_by_gep(ctx, pobj, obj_name), value).unwrap();
}
}
impl<'ctx, Value> From<StructField<'ctx, Value>> for (&'static str, BasicTypeEnum<'ctx>)
where
Value: BasicValue<'ctx> + TryFrom<BasicValueEnum<'ctx>, Error = ()>,
{
fn from(value: StructField<'ctx, Value>) -> Self {
(value.name, value.ty)
}
}
/// A counter that tracks the next index of a field using a monotonically increasing counter.
#[derive(Default, Debug, PartialEq, Eq, Clone, Copy)]
pub struct FieldIndexCounter(u32);
impl FieldIndexCounter {
/// Increments the number stored by this counter, returning the previous value.
///
/// Functionally equivalent to `i++` in C-based languages.
pub fn increment(&mut self) -> u32 {
let v = self.0;
self.0 += 1;
v
}
}
type FieldTypeVerifier<'ctx> = dyn Fn(BasicTypeEnum<'ctx>) -> Result<(), String>;
/// Checks whether [`llvm_ty`][StructType] contains the fields described by the given
/// [`StructFields`] instance.
///
/// By default, this function will compare the type of each field in `expected_fields` against
/// `llvm_ty`. To override this behavior for individual fields, pass in overrides to
/// `custom_verifiers`, which will use the specified verifier when a field with the matching field
/// name is being checked.
pub(super) fn check_struct_type_matches_fields<'ctx>(
expected_fields: impl StructFields<'ctx>,
llvm_ty: StructType<'ctx>,
ty_name: &'static str,
custom_verifiers: &[(&str, &FieldTypeVerifier<'ctx>)],
) -> Result<(), String> {
let expected_fields = expected_fields.to_vec();
if llvm_ty.count_fields() != u32::try_from(expected_fields.len()).unwrap() {
return Err(format!(
"Expected {} fields in `{ty_name}`, got {}",
expected_fields.len(),
llvm_ty.count_fields(),
));
}
expected_fields
.into_iter()
.enumerate()
.map(|(i, (field_name, expected_ty))| {
(field_name, expected_ty, llvm_ty.get_field_type_at_index(i as u32).unwrap())
})
.try_for_each(|(field_name, expected_ty, actual_ty)| {
if let Some((_, verifier)) =
custom_verifiers.iter().find(|verifier| verifier.0 == field_name)
{
verifier(actual_ty)
} else if expected_ty == actual_ty {
Ok(())
} else {
Err(format!("Expected {expected_ty} for `{ty_name}.{field_name}`, got {actual_ty}"))
}
})?;
Ok(())
}

View File

@ -0,0 +1,3 @@
pub use slice::*;
mod slice;

View File

@ -0,0 +1,254 @@
use inkwell::{
context::{AsContextRef, Context, ContextRef},
types::{AnyTypeEnum, BasicType, BasicTypeEnum, IntType, PointerType},
values::IntValue,
AddressSpace,
};
use itertools::Itertools;
use nac3core_derive::StructFields;
use crate::codegen::{
types::{
structure::{
check_struct_type_matches_fields, FieldIndexCounter, StructField, StructFields,
},
ProxyType,
},
values::{utils::SliceValue, ArraySliceValue, ProxyValue},
CodeGenContext, CodeGenerator,
};
#[derive(Debug, PartialEq, Eq, Clone, Copy)]
pub struct SliceType<'ctx> {
ty: PointerType<'ctx>,
int_ty: IntType<'ctx>,
llvm_usize: IntType<'ctx>,
}
#[derive(PartialEq, Eq, Clone, Copy, StructFields)]
pub struct SliceFields<'ctx> {
#[value_type(bool_type())]
pub start_defined: StructField<'ctx, IntValue<'ctx>>,
#[value_type(usize)]
pub start: StructField<'ctx, IntValue<'ctx>>,
#[value_type(bool_type())]
pub stop_defined: StructField<'ctx, IntValue<'ctx>>,
#[value_type(usize)]
pub stop: StructField<'ctx, IntValue<'ctx>>,
#[value_type(bool_type())]
pub step_defined: StructField<'ctx, IntValue<'ctx>>,
#[value_type(usize)]
pub step: StructField<'ctx, IntValue<'ctx>>,
}
impl<'ctx> SliceFields<'ctx> {
/// Creates a new instance of [`SliceFields`] with a custom integer type for its range values.
#[must_use]
pub fn new_sized(ctx: &impl AsContextRef<'ctx>, int_ty: IntType<'ctx>) -> Self {
let ctx = unsafe { ContextRef::new(ctx.as_ctx_ref()) };
let mut counter = FieldIndexCounter::default();
SliceFields {
start_defined: StructField::create(&mut counter, "start_defined", ctx.bool_type()),
start: StructField::create(&mut counter, "start", int_ty),
stop_defined: StructField::create(&mut counter, "stop_defined", ctx.bool_type()),
stop: StructField::create(&mut counter, "stop", int_ty),
step_defined: StructField::create(&mut counter, "step_defined", ctx.bool_type()),
step: StructField::create(&mut counter, "step", int_ty),
}
}
}
impl<'ctx> SliceType<'ctx> {
/// Checks whether `llvm_ty` represents a `slice` type, returning [Err] if it does not.
pub fn is_representable(
llvm_ty: PointerType<'ctx>,
llvm_usize: IntType<'ctx>,
) -> Result<(), String> {
let ctx = llvm_ty.get_context();
let fields = SliceFields::new(ctx, llvm_usize);
let llvm_ty = llvm_ty.get_element_type();
let AnyTypeEnum::StructType(llvm_ty) = llvm_ty else {
return Err(format!("Expected struct type for `Slice` type, got {llvm_ty}"));
};
check_struct_type_matches_fields(
fields,
llvm_ty,
"Slice",
&[
(fields.start.name(), &|ty| {
if ty.is_int_type() {
Ok(())
} else {
Err(format!("Expected int type for `Slice.start`, got {ty}"))
}
}),
(fields.stop.name(), &|ty| {
if ty.is_int_type() {
Ok(())
} else {
Err(format!("Expected int type for `Slice.stop`, got {ty}"))
}
}),
(fields.step.name(), &|ty| {
if ty.is_int_type() {
Ok(())
} else {
Err(format!("Expected int type for `Slice.step`, got {ty}"))
}
}),
],
)
}
// TODO: Move this into e.g. StructProxyType
#[must_use]
pub fn get_fields(&self) -> SliceFields<'ctx> {
SliceFields::new_sized(&self.int_ty.get_context(), self.int_ty)
}
/// Creates an LLVM type corresponding to the expected structure of a `Slice`.
#[must_use]
fn llvm_type(ctx: &'ctx Context, int_ty: IntType<'ctx>) -> PointerType<'ctx> {
let field_tys = SliceFields::new_sized(&int_ty.get_context(), int_ty)
.into_iter()
.map(|field| field.1)
.collect_vec();
ctx.struct_type(&field_tys, false).ptr_type(AddressSpace::default())
}
/// Creates an instance of [`SliceType`] with `int_ty` as its backing integer type.
#[must_use]
pub fn new(ctx: &'ctx Context, int_ty: IntType<'ctx>, llvm_usize: IntType<'ctx>) -> Self {
let llvm_ty = Self::llvm_type(ctx, int_ty);
Self { ty: llvm_ty, int_ty, llvm_usize }
}
/// Creates an instance of [`SliceType`] with `usize` as its backing integer type.
#[must_use]
pub fn new_usize<G: CodeGenerator + ?Sized>(generator: &G, ctx: &'ctx Context) -> Self {
let llvm_usize = generator.get_size_type(ctx);
Self::new(ctx, llvm_usize, llvm_usize)
}
/// Creates an [`SliceType`] from a [`PointerType`] representing a `slice`.
#[must_use]
pub fn from_type(
ptr_ty: PointerType<'ctx>,
int_ty: IntType<'ctx>,
llvm_usize: IntType<'ctx>,
) -> Self {
debug_assert!(Self::is_representable(ptr_ty, int_ty).is_ok());
Self { ty: ptr_ty, int_ty, llvm_usize }
}
#[must_use]
pub fn element_type(&self) -> IntType<'ctx> {
self.int_ty
}
/// Allocates an instance of [`ContiguousNDArrayValue`] as if by calling `alloca` on the base type.
#[must_use]
pub fn alloca<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
name: Option<&'ctx str>,
) -> <Self as ProxyType<'ctx>>::Value {
<Self as ProxyType<'ctx>>::Value::from_pointer_value(
self.raw_alloca(generator, ctx, name),
self.int_ty,
self.llvm_usize,
name,
)
}
/// Converts an existing value into a [`ContiguousNDArrayValue`].
#[must_use]
pub fn map_value(
&self,
value: <<Self as ProxyType<'ctx>>::Value as ProxyValue<'ctx>>::Base,
name: Option<&'ctx str>,
) -> <Self as ProxyType<'ctx>>::Value {
<Self as ProxyType<'ctx>>::Value::from_pointer_value(
value,
self.int_ty,
self.llvm_usize,
name,
)
}
}
impl<'ctx> ProxyType<'ctx> for SliceType<'ctx> {
type Base = PointerType<'ctx>;
type Value = SliceValue<'ctx>;
fn is_type<G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &'ctx Context,
llvm_ty: impl BasicType<'ctx>,
) -> Result<(), String> {
if let BasicTypeEnum::PointerType(ty) = llvm_ty.as_basic_type_enum() {
<Self as ProxyType<'ctx>>::is_representable(generator, ctx, ty)
} else {
Err(format!("Expected pointer type, got {llvm_ty:?}"))
}
}
fn is_representable<G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &'ctx Context,
llvm_ty: Self::Base,
) -> Result<(), String> {
Self::is_representable(llvm_ty, generator.get_size_type(ctx))
}
fn raw_alloca<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
name: Option<&'ctx str>,
) -> <Self::Value as ProxyValue<'ctx>>::Base {
generator
.gen_var_alloc(
ctx,
self.as_base_type().get_element_type().into_struct_type().into(),
name,
)
.unwrap()
}
fn array_alloca<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
size: IntValue<'ctx>,
name: Option<&'ctx str>,
) -> ArraySliceValue<'ctx> {
generator
.gen_array_var_alloc(
ctx,
self.as_base_type().get_element_type().into_struct_type().into(),
size,
name,
)
.unwrap()
}
fn as_base_type(&self) -> Self::Base {
self.ty
}
}
impl<'ctx> From<SliceType<'ctx>> for PointerType<'ctx> {
fn from(value: SliceType<'ctx>) -> Self {
value.as_base_type()
}
}

View File

@ -0,0 +1,426 @@
use inkwell::{
types::AnyTypeEnum,
values::{BasicValueEnum, IntValue, PointerValue},
IntPredicate,
};
use crate::codegen::{CodeGenContext, CodeGenerator};
/// An LLVM value that is array-like, i.e. it contains a contiguous, sequenced collection of
/// elements.
pub trait ArrayLikeValue<'ctx> {
/// Returns the element type of this array-like value.
fn element_type<G: CodeGenerator + ?Sized>(
&self,
ctx: &CodeGenContext<'ctx, '_>,
generator: &G,
) -> AnyTypeEnum<'ctx>;
/// Returns the base pointer to the array.
fn base_ptr<G: CodeGenerator + ?Sized>(
&self,
ctx: &CodeGenContext<'ctx, '_>,
generator: &G,
) -> PointerValue<'ctx>;
/// Returns the size of this array-like value.
fn size<G: CodeGenerator + ?Sized>(
&self,
ctx: &CodeGenContext<'ctx, '_>,
generator: &G,
) -> IntValue<'ctx>;
/// Returns a [`ArraySliceValue`] representing this value.
fn as_slice_value<G: CodeGenerator + ?Sized>(
&self,
ctx: &CodeGenContext<'ctx, '_>,
generator: &G,
) -> ArraySliceValue<'ctx> {
ArraySliceValue::from_ptr_val(
self.base_ptr(ctx, generator),
self.size(ctx, generator),
None,
)
}
}
/// An array-like value that can be indexed by memory offset.
pub trait ArrayLikeIndexer<'ctx, Index = IntValue<'ctx>>: ArrayLikeValue<'ctx> {
/// # Safety
///
/// This function should be called with a valid index.
unsafe fn ptr_offset_unchecked<G: CodeGenerator + ?Sized>(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
idx: &Index,
name: Option<&str>,
) -> PointerValue<'ctx>;
/// Returns the pointer to the data at the `idx`-th index.
fn ptr_offset<G: CodeGenerator + ?Sized>(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
idx: &Index,
name: Option<&str>,
) -> PointerValue<'ctx>;
}
/// An array-like value that can have its array elements accessed as a [`BasicValueEnum`].
pub trait UntypedArrayLikeAccessor<'ctx, Index = IntValue<'ctx>>:
ArrayLikeIndexer<'ctx, Index>
{
/// # Safety
///
/// This function should be called with a valid index.
unsafe fn get_unchecked<G: CodeGenerator + ?Sized>(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
idx: &Index,
name: Option<&str>,
) -> BasicValueEnum<'ctx> {
let ptr = unsafe { self.ptr_offset_unchecked(ctx, generator, idx, name) };
ctx.builder.build_load(ptr, name.unwrap_or_default()).unwrap()
}
/// Returns the data at the `idx`-th index.
fn get<G: CodeGenerator + ?Sized>(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
idx: &Index,
name: Option<&str>,
) -> BasicValueEnum<'ctx> {
let ptr = self.ptr_offset(ctx, generator, idx, name);
ctx.builder.build_load(ptr, name.unwrap_or_default()).unwrap()
}
}
/// An array-like value that can have its array elements mutated as a [`BasicValueEnum`].
pub trait UntypedArrayLikeMutator<'ctx, Index = IntValue<'ctx>>:
ArrayLikeIndexer<'ctx, Index>
{
/// # Safety
///
/// This function should be called with a valid index.
unsafe fn set_unchecked<G: CodeGenerator + ?Sized>(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
idx: &Index,
value: BasicValueEnum<'ctx>,
) {
let ptr = unsafe { self.ptr_offset_unchecked(ctx, generator, idx, None) };
ctx.builder.build_store(ptr, value).unwrap();
}
/// Sets the data at the `idx`-th index.
fn set<G: CodeGenerator + ?Sized>(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
idx: &Index,
value: BasicValueEnum<'ctx>,
) {
let ptr = self.ptr_offset(ctx, generator, idx, None);
ctx.builder.build_store(ptr, value).unwrap();
}
}
/// An array-like value that can have its array elements accessed as an arbitrary type `T`.
pub trait TypedArrayLikeAccessor<'ctx, T, Index = IntValue<'ctx>>:
UntypedArrayLikeAccessor<'ctx, Index>
{
/// Casts an element from [`BasicValueEnum`] into `T`.
fn downcast_to_type(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
value: BasicValueEnum<'ctx>,
) -> T;
/// # Safety
///
/// This function should be called with a valid index.
unsafe fn get_typed_unchecked<G: CodeGenerator + ?Sized>(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
idx: &Index,
name: Option<&str>,
) -> T {
let value = unsafe { self.get_unchecked(ctx, generator, idx, name) };
self.downcast_to_type(ctx, value)
}
/// Returns the data at the `idx`-th index.
fn get_typed<G: CodeGenerator + ?Sized>(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
idx: &Index,
name: Option<&str>,
) -> T {
let value = self.get(ctx, generator, idx, name);
self.downcast_to_type(ctx, value)
}
}
/// An array-like value that can have its array elements mutated as an arbitrary type `T`.
pub trait TypedArrayLikeMutator<'ctx, T, Index = IntValue<'ctx>>:
UntypedArrayLikeMutator<'ctx, Index>
{
/// Casts an element from T into [`BasicValueEnum`].
fn upcast_from_type(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
value: T,
) -> BasicValueEnum<'ctx>;
/// # Safety
///
/// This function should be called with a valid index.
unsafe fn set_typed_unchecked<G: CodeGenerator + ?Sized>(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
idx: &Index,
value: T,
) {
let value = self.upcast_from_type(ctx, value);
unsafe { self.set_unchecked(ctx, generator, idx, value) }
}
/// Sets the data at the `idx`-th index.
fn set_typed<G: CodeGenerator + ?Sized>(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
idx: &Index,
value: T,
) {
let value = self.upcast_from_type(ctx, value);
self.set(ctx, generator, idx, value);
}
}
/// Type alias for a function that casts a [`BasicValueEnum`] into a `T`.
type ValueDowncastFn<'ctx, T> =
Box<dyn Fn(&mut CodeGenContext<'ctx, '_>, BasicValueEnum<'ctx>) -> T + 'ctx>;
/// Type alias for a function that casts a `T` into a [`BasicValueEnum`].
type ValueUpcastFn<'ctx, T> = Box<dyn Fn(&mut CodeGenContext<'ctx, '_>, T) -> BasicValueEnum<'ctx>>;
/// An adapter for constraining untyped array values as typed values.
pub struct TypedArrayLikeAdapter<'ctx, T, Adapted: ArrayLikeValue<'ctx> = ArraySliceValue<'ctx>> {
adapted: Adapted,
downcast_fn: ValueDowncastFn<'ctx, T>,
upcast_fn: ValueUpcastFn<'ctx, T>,
}
impl<'ctx, T, Adapted> TypedArrayLikeAdapter<'ctx, T, Adapted>
where
Adapted: ArrayLikeValue<'ctx>,
{
/// Creates a [`TypedArrayLikeAdapter`].
///
/// * `adapted` - The value to be adapted.
/// * `downcast_fn` - The function converting a [`BasicValueEnum`] into a `T`.
/// * `upcast_fn` - The function converting a T into a [`BasicValueEnum`].
pub fn from(
adapted: Adapted,
downcast_fn: ValueDowncastFn<'ctx, T>,
upcast_fn: ValueUpcastFn<'ctx, T>,
) -> Self {
TypedArrayLikeAdapter { adapted, downcast_fn, upcast_fn }
}
}
impl<'ctx, T, Adapted> ArrayLikeValue<'ctx> for TypedArrayLikeAdapter<'ctx, T, Adapted>
where
Adapted: ArrayLikeValue<'ctx>,
{
fn element_type<G: CodeGenerator + ?Sized>(
&self,
ctx: &CodeGenContext<'ctx, '_>,
generator: &G,
) -> AnyTypeEnum<'ctx> {
self.adapted.element_type(ctx, generator)
}
fn base_ptr<G: CodeGenerator + ?Sized>(
&self,
ctx: &CodeGenContext<'ctx, '_>,
generator: &G,
) -> PointerValue<'ctx> {
self.adapted.base_ptr(ctx, generator)
}
fn size<G: CodeGenerator + ?Sized>(
&self,
ctx: &CodeGenContext<'ctx, '_>,
generator: &G,
) -> IntValue<'ctx> {
self.adapted.size(ctx, generator)
}
}
impl<'ctx, T, Index, Adapted> ArrayLikeIndexer<'ctx, Index>
for TypedArrayLikeAdapter<'ctx, T, Adapted>
where
Adapted: ArrayLikeIndexer<'ctx, Index>,
{
unsafe fn ptr_offset_unchecked<G: CodeGenerator + ?Sized>(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
idx: &Index,
name: Option<&str>,
) -> PointerValue<'ctx> {
unsafe { self.adapted.ptr_offset_unchecked(ctx, generator, idx, name) }
}
fn ptr_offset<G: CodeGenerator + ?Sized>(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
idx: &Index,
name: Option<&str>,
) -> PointerValue<'ctx> {
self.adapted.ptr_offset(ctx, generator, idx, name)
}
}
impl<'ctx, T, Index, Adapted> UntypedArrayLikeAccessor<'ctx, Index>
for TypedArrayLikeAdapter<'ctx, T, Adapted>
where
Adapted: UntypedArrayLikeAccessor<'ctx, Index>,
{
}
impl<'ctx, T, Index, Adapted> UntypedArrayLikeMutator<'ctx, Index>
for TypedArrayLikeAdapter<'ctx, T, Adapted>
where
Adapted: UntypedArrayLikeMutator<'ctx, Index>,
{
}
impl<'ctx, T, Index, Adapted> TypedArrayLikeAccessor<'ctx, T, Index>
for TypedArrayLikeAdapter<'ctx, T, Adapted>
where
Adapted: UntypedArrayLikeAccessor<'ctx, Index>,
{
fn downcast_to_type(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
value: BasicValueEnum<'ctx>,
) -> T {
(self.downcast_fn)(ctx, value)
}
}
impl<'ctx, T, Index, Adapted> TypedArrayLikeMutator<'ctx, T, Index>
for TypedArrayLikeAdapter<'ctx, T, Adapted>
where
Adapted: UntypedArrayLikeMutator<'ctx, Index>,
{
fn upcast_from_type(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
value: T,
) -> BasicValueEnum<'ctx> {
(self.upcast_fn)(ctx, value)
}
}
/// An LLVM value representing an array slice, consisting of a pointer to the data and the size of
/// the slice.
#[derive(Copy, Clone)]
pub struct ArraySliceValue<'ctx>(PointerValue<'ctx>, IntValue<'ctx>, Option<&'ctx str>);
impl<'ctx> ArraySliceValue<'ctx> {
/// Creates an [`ArraySliceValue`] from a [`PointerValue`] and its size.
#[must_use]
pub fn from_ptr_val(
ptr: PointerValue<'ctx>,
size: IntValue<'ctx>,
name: Option<&'ctx str>,
) -> Self {
ArraySliceValue(ptr, size, name)
}
}
impl<'ctx> From<ArraySliceValue<'ctx>> for PointerValue<'ctx> {
fn from(value: ArraySliceValue<'ctx>) -> Self {
value.0
}
}
impl<'ctx> ArrayLikeValue<'ctx> for ArraySliceValue<'ctx> {
fn element_type<G: CodeGenerator + ?Sized>(
&self,
_: &CodeGenContext<'ctx, '_>,
_: &G,
) -> AnyTypeEnum<'ctx> {
self.0.get_type().get_element_type()
}
fn base_ptr<G: CodeGenerator + ?Sized>(
&self,
_: &CodeGenContext<'ctx, '_>,
_: &G,
) -> PointerValue<'ctx> {
self.0
}
fn size<G: CodeGenerator + ?Sized>(
&self,
_: &CodeGenContext<'ctx, '_>,
_: &G,
) -> IntValue<'ctx> {
self.1
}
}
impl<'ctx> ArrayLikeIndexer<'ctx> for ArraySliceValue<'ctx> {
unsafe fn ptr_offset_unchecked<G: CodeGenerator + ?Sized>(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
idx: &IntValue<'ctx>,
name: Option<&str>,
) -> PointerValue<'ctx> {
let var_name = name.map(|v| format!("{v}.addr")).unwrap_or_default();
unsafe {
ctx.builder
.build_in_bounds_gep(self.base_ptr(ctx, generator), &[*idx], var_name.as_str())
.unwrap()
}
}
fn ptr_offset<G: CodeGenerator + ?Sized>(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
idx: &IntValue<'ctx>,
name: Option<&str>,
) -> PointerValue<'ctx> {
debug_assert_eq!(idx.get_type(), generator.get_size_type(ctx.ctx));
let size = self.size(ctx, generator);
let in_range = ctx.builder.build_int_compare(IntPredicate::ULT, *idx, size, "").unwrap();
ctx.make_assert(
generator,
in_range,
"0:IndexError",
"list index out of range",
[None, None, None],
ctx.current_loc,
);
unsafe { self.ptr_offset_unchecked(ctx, generator, idx, name) }
}
}
impl<'ctx> UntypedArrayLikeAccessor<'ctx> for ArraySliceValue<'ctx> {}
impl<'ctx> UntypedArrayLikeMutator<'ctx> for ArraySliceValue<'ctx> {}

View File

@ -0,0 +1,241 @@
use inkwell::{
types::{AnyTypeEnum, BasicType, BasicTypeEnum, IntType},
values::{BasicValueEnum, IntValue, PointerValue},
AddressSpace, IntPredicate,
};
use super::{
ArrayLikeIndexer, ArrayLikeValue, ProxyValue, UntypedArrayLikeAccessor, UntypedArrayLikeMutator,
};
use crate::codegen::{
types::ListType,
{CodeGenContext, CodeGenerator},
};
/// Proxy type for accessing a `list` value in LLVM.
#[derive(Copy, Clone)]
pub struct ListValue<'ctx> {
value: PointerValue<'ctx>,
llvm_usize: IntType<'ctx>,
name: Option<&'ctx str>,
}
impl<'ctx> ListValue<'ctx> {
/// Checks whether `value` is an instance of `list`, returning [Err] if `value` is not an
/// instance.
pub fn is_representable(
value: PointerValue<'ctx>,
llvm_usize: IntType<'ctx>,
) -> Result<(), String> {
ListType::is_representable(value.get_type(), llvm_usize)
}
/// Creates an [`ListValue`] from a [`PointerValue`].
#[must_use]
pub fn from_pointer_value(
ptr: PointerValue<'ctx>,
llvm_usize: IntType<'ctx>,
name: Option<&'ctx str>,
) -> Self {
debug_assert!(Self::is_representable(ptr, llvm_usize).is_ok());
ListValue { value: ptr, llvm_usize, name }
}
/// Returns the double-indirection pointer to the `data` array, as if by calling `getelementptr`
/// on the field.
fn pptr_to_data(&self, ctx: &CodeGenContext<'ctx, '_>) -> PointerValue<'ctx> {
let llvm_i32 = ctx.ctx.i32_type();
let var_name = self.name.map(|v| format!("{v}.data.addr")).unwrap_or_default();
unsafe {
ctx.builder
.build_in_bounds_gep(
self.as_base_value(),
&[llvm_i32.const_zero(), llvm_i32.const_zero()],
var_name.as_str(),
)
.unwrap()
}
}
/// Returns the pointer to the field storing the size of this `list`.
fn ptr_to_size(&self, ctx: &CodeGenContext<'ctx, '_>) -> PointerValue<'ctx> {
let llvm_i32 = ctx.ctx.i32_type();
let var_name = self.name.map(|v| format!("{v}.size.addr")).unwrap_or_default();
unsafe {
ctx.builder
.build_in_bounds_gep(
self.as_base_value(),
&[llvm_i32.const_zero(), llvm_i32.const_int(1, true)],
var_name.as_str(),
)
.unwrap()
}
}
/// Stores the array of data elements `data` into this instance.
fn store_data(&self, ctx: &CodeGenContext<'ctx, '_>, data: PointerValue<'ctx>) {
ctx.builder.build_store(self.pptr_to_data(ctx), data).unwrap();
}
/// Convenience method for creating a new array storing data elements with the given element
/// type `elem_ty` and `size`.
///
/// If `size` is [None], the size stored in the field of this instance is used instead.
pub fn create_data(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
elem_ty: BasicTypeEnum<'ctx>,
size: Option<IntValue<'ctx>>,
) {
let size = size.unwrap_or_else(|| self.load_size(ctx, None));
let data = ctx
.builder
.build_select(
ctx.builder
.build_int_compare(IntPredicate::NE, size, self.llvm_usize.const_zero(), "")
.unwrap(),
ctx.builder.build_array_alloca(elem_ty, size, "").unwrap(),
elem_ty.ptr_type(AddressSpace::default()).const_zero(),
"",
)
.map(BasicValueEnum::into_pointer_value)
.unwrap();
self.store_data(ctx, data);
}
/// Returns the double-indirection pointer to the `data` array, as if by calling `getelementptr`
/// on the field.
#[must_use]
pub fn data(&self) -> ListDataProxy<'ctx, '_> {
ListDataProxy(self)
}
/// Stores the `size` of this `list` into this instance.
pub fn store_size<G: CodeGenerator + ?Sized>(
&self,
ctx: &CodeGenContext<'ctx, '_>,
generator: &G,
size: IntValue<'ctx>,
) {
debug_assert_eq!(size.get_type(), generator.get_size_type(ctx.ctx));
let psize = self.ptr_to_size(ctx);
ctx.builder.build_store(psize, size).unwrap();
}
/// Returns the size of this `list` as a value.
pub fn load_size(&self, ctx: &CodeGenContext<'ctx, '_>, name: Option<&str>) -> IntValue<'ctx> {
let psize = self.ptr_to_size(ctx);
let var_name = name
.map(ToString::to_string)
.or_else(|| self.name.map(|v| format!("{v}.size")))
.unwrap_or_default();
ctx.builder
.build_load(psize, var_name.as_str())
.map(BasicValueEnum::into_int_value)
.unwrap()
}
}
impl<'ctx> ProxyValue<'ctx> for ListValue<'ctx> {
type Base = PointerValue<'ctx>;
type Type = ListType<'ctx>;
fn get_type(&self) -> Self::Type {
ListType::from_type(self.as_base_value().get_type(), self.llvm_usize)
}
fn as_base_value(&self) -> Self::Base {
self.value
}
}
impl<'ctx> From<ListValue<'ctx>> for PointerValue<'ctx> {
fn from(value: ListValue<'ctx>) -> Self {
value.as_base_value()
}
}
/// Proxy type for accessing the `data` array of an `list` instance in LLVM.
#[derive(Copy, Clone)]
pub struct ListDataProxy<'ctx, 'a>(&'a ListValue<'ctx>);
impl<'ctx> ArrayLikeValue<'ctx> for ListDataProxy<'ctx, '_> {
fn element_type<G: CodeGenerator + ?Sized>(
&self,
_: &CodeGenContext<'ctx, '_>,
_: &G,
) -> AnyTypeEnum<'ctx> {
self.0.value.get_type().get_element_type()
}
fn base_ptr<G: CodeGenerator + ?Sized>(
&self,
ctx: &CodeGenContext<'ctx, '_>,
_: &G,
) -> PointerValue<'ctx> {
let var_name = self.0.name.map(|v| format!("{v}.data")).unwrap_or_default();
ctx.builder
.build_load(self.0.pptr_to_data(ctx), var_name.as_str())
.map(BasicValueEnum::into_pointer_value)
.unwrap()
}
fn size<G: CodeGenerator + ?Sized>(
&self,
ctx: &CodeGenContext<'ctx, '_>,
_: &G,
) -> IntValue<'ctx> {
self.0.load_size(ctx, None)
}
}
impl<'ctx> ArrayLikeIndexer<'ctx> for ListDataProxy<'ctx, '_> {
unsafe fn ptr_offset_unchecked<G: CodeGenerator + ?Sized>(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
idx: &IntValue<'ctx>,
name: Option<&str>,
) -> PointerValue<'ctx> {
let var_name = name.map(|v| format!("{v}.addr")).unwrap_or_default();
unsafe {
ctx.builder
.build_in_bounds_gep(self.base_ptr(ctx, generator), &[*idx], var_name.as_str())
.unwrap()
}
}
fn ptr_offset<G: CodeGenerator + ?Sized>(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
idx: &IntValue<'ctx>,
name: Option<&str>,
) -> PointerValue<'ctx> {
debug_assert_eq!(idx.get_type(), generator.get_size_type(ctx.ctx));
let size = self.size(ctx, generator);
let in_range = ctx.builder.build_int_compare(IntPredicate::ULT, *idx, size, "").unwrap();
ctx.make_assert(
generator,
in_range,
"0:IndexError",
"list index out of range",
[None, None, None],
ctx.current_loc,
);
unsafe { self.ptr_offset_unchecked(ctx, generator, idx, name) }
}
}
impl<'ctx> UntypedArrayLikeAccessor<'ctx> for ListDataProxy<'ctx, '_> {}
impl<'ctx> UntypedArrayLikeMutator<'ctx> for ListDataProxy<'ctx, '_> {}

View File

@ -0,0 +1,47 @@
use inkwell::{context::Context, values::BasicValue};
use super::types::ProxyType;
use crate::codegen::CodeGenerator;
pub use array::*;
pub use list::*;
pub use range::*;
mod array;
mod list;
pub mod ndarray;
mod range;
pub mod utils;
/// A LLVM type that is used to represent a non-primitive value in NAC3.
pub trait ProxyValue<'ctx>: Into<Self::Base> {
/// The type of LLVM values represented by this instance. This is usually the
/// [LLVM pointer type][PointerValue].
type Base: BasicValue<'ctx>;
/// The type of this value.
type Type: ProxyType<'ctx, Value = Self>;
/// Checks whether `value` can be represented by this [`ProxyValue`].
fn is_instance<G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &'ctx Context,
value: impl BasicValue<'ctx>,
) -> Result<(), String> {
Self::Type::is_type(generator, ctx, value.as_basic_value_enum().get_type())
}
/// Checks whether `value` can be represented by this [`ProxyValue`].
fn is_representable<G: CodeGenerator + ?Sized>(
generator: &G,
ctx: &'ctx Context,
value: Self::Base,
) -> Result<(), String> {
Self::is_instance(generator, ctx, value.as_basic_value_enum())
}
/// Returns the [type][ProxyType] of this value.
fn get_type(&self) -> Self::Type;
/// Returns the [base value][Self::Base] of this proxy.
fn as_base_value(&self) -> Self::Base;
}

View File

@ -0,0 +1,202 @@
use inkwell::{
types::{BasicType, BasicTypeEnum, IntType},
values::{IntValue, PointerValue},
AddressSpace,
};
use super::{ArrayLikeValue, NDArrayValue, ProxyValue};
use crate::codegen::{
stmt::gen_if_callback,
types::{
ndarray::{ContiguousNDArrayType, NDArrayType},
structure::StructField,
},
CodeGenContext, CodeGenerator,
};
#[derive(Copy, Clone)]
pub struct ContiguousNDArrayValue<'ctx> {
value: PointerValue<'ctx>,
item: BasicTypeEnum<'ctx>,
llvm_usize: IntType<'ctx>,
name: Option<&'ctx str>,
}
impl<'ctx> ContiguousNDArrayValue<'ctx> {
/// Checks whether `value` is an instance of `ContiguousNDArray`, returning [Err] if `value` is
/// not an instance.
pub fn is_representable(
value: PointerValue<'ctx>,
llvm_usize: IntType<'ctx>,
) -> Result<(), String> {
<Self as ProxyValue<'ctx>>::Type::is_representable(value.get_type(), llvm_usize)
}
/// Creates an [`ContiguousNDArrayValue`] from a [`PointerValue`].
#[must_use]
pub fn from_pointer_value(
ptr: PointerValue<'ctx>,
dtype: BasicTypeEnum<'ctx>,
llvm_usize: IntType<'ctx>,
name: Option<&'ctx str>,
) -> Self {
debug_assert!(Self::is_representable(ptr, llvm_usize).is_ok());
Self { value: ptr, item: dtype, llvm_usize, name }
}
fn ndims_field(&self) -> StructField<'ctx, IntValue<'ctx>> {
self.get_type().get_fields().ndims
}
pub fn store_ndims(&self, ctx: &CodeGenContext<'ctx, '_>, value: IntValue<'ctx>) {
self.ndims_field().set(ctx, self.as_base_value(), value, self.name);
}
fn shape_field(&self) -> StructField<'ctx, PointerValue<'ctx>> {
self.get_type().get_fields().shape
}
pub fn store_shape(&self, ctx: &CodeGenContext<'ctx, '_>, value: PointerValue<'ctx>) {
self.shape_field().set(ctx, self.as_base_value(), value, self.name);
}
pub fn load_shape(&self, ctx: &CodeGenContext<'ctx, '_>) -> PointerValue<'ctx> {
self.shape_field().get(ctx, self.value, self.name)
}
fn data_field(&self) -> StructField<'ctx, PointerValue<'ctx>> {
self.get_type().get_fields().data
}
pub fn store_data(&self, ctx: &CodeGenContext<'ctx, '_>, value: PointerValue<'ctx>) {
self.data_field().set(ctx, self.as_base_value(), value, self.name);
}
pub fn load_data(&self, ctx: &CodeGenContext<'ctx, '_>) -> PointerValue<'ctx> {
self.data_field().get(ctx, self.value, self.name)
}
}
impl<'ctx> ProxyValue<'ctx> for ContiguousNDArrayValue<'ctx> {
type Base = PointerValue<'ctx>;
type Type = ContiguousNDArrayType<'ctx>;
fn get_type(&self) -> Self::Type {
<Self as ProxyValue<'ctx>>::Type::from_type(
self.as_base_value().get_type(),
self.item,
self.llvm_usize,
)
}
fn as_base_value(&self) -> Self::Base {
self.value
}
}
impl<'ctx> From<ContiguousNDArrayValue<'ctx>> for PointerValue<'ctx> {
fn from(value: ContiguousNDArrayValue<'ctx>) -> Self {
value.as_base_value()
}
}
impl<'ctx> NDArrayValue<'ctx> {
/// Create a [`ContiguousNDArrayValue`] from the contents of this ndarray.
///
/// This function may or may not be expensive depending on if this ndarray has contiguous data.
///
/// If this ndarray is not C-contiguous, this function will allocate memory on the stack for the
/// `data` field of the returned [`ContiguousNDArrayValue`] and copy contents of this ndarray to
/// there.
///
/// If this ndarray is C-contiguous, contents of this ndarray will not be copied. The created
/// [`ContiguousNDArrayValue`] will share memory with this ndarray.
pub fn make_contiguous_ndarray<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
) -> ContiguousNDArrayValue<'ctx> {
let result = ContiguousNDArrayType::new(generator, ctx.ctx, self.dtype)
.alloca(generator, ctx, self.name);
// Set ndims and shape.
let ndims = self
.ndims
.map_or_else(|| self.load_ndims(ctx), |ndims| self.llvm_usize.const_int(ndims, false));
result.store_ndims(ctx, ndims);
let shape = self.shape();
result.store_shape(ctx, shape.base_ptr(ctx, generator));
gen_if_callback(
generator,
ctx,
|generator, ctx| Ok(self.is_c_contiguous(generator, ctx)),
|_, ctx| {
// This ndarray is contiguous.
let data = self.data_field(ctx).get(ctx, self.as_base_value(), self.name);
let data = ctx
.builder
.build_pointer_cast(data, result.item.ptr_type(AddressSpace::default()), "")
.unwrap();
result.store_data(ctx, data);
Ok(())
},
|generator, ctx| {
// This ndarray is not contiguous. Do a full-copy on `data`. `make_copy` produces an
// ndarray with contiguous `data`.
let copied_ndarray = self.make_copy(generator, ctx);
let data = copied_ndarray.data().base_ptr(ctx, generator);
let data = ctx
.builder
.build_pointer_cast(data, result.item.ptr_type(AddressSpace::default()), "")
.unwrap();
result.store_data(ctx, data);
Ok(())
},
)
.unwrap();
result
}
/// Create an [`NDArrayValue`] from a [`ContiguousNDArrayValue`].
///
/// The operation is cheap. The newly created [`NDArrayValue`] will share the same memory as the
/// [`ContiguousNDArrayValue`].
///
/// `ndims` has to be provided as [`NDArrayValue`] requires a statically known `ndims` value,
/// despite the fact that the information should be contained within the
/// [`ContiguousNDArrayValue`].
pub fn from_contiguous_ndarray<G: CodeGenerator + ?Sized>(
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
carray: ContiguousNDArrayValue<'ctx>,
ndims: u64,
) -> Self {
// TODO: Debug assert `ndims == carray.ndims` to catch bugs.
// Allocate the resulting ndarray.
let ndarray = NDArrayType::new(generator, ctx.ctx, carray.item, Some(ndims))
.construct_uninitialized(generator, ctx, carray.name);
// Copy shape and update strides
let shape = carray.load_shape(ctx);
ndarray.copy_shape_from_array(generator, ctx, shape);
ndarray.set_strides_contiguous(generator, ctx);
// Share data
let data = carray.load_data(ctx);
ndarray.store_data(
ctx,
ctx.builder
.build_pointer_cast(data, ctx.ctx.i8_type().ptr_type(AddressSpace::default()), "")
.unwrap(),
);
ndarray
}
}

View File

@ -0,0 +1,262 @@
use inkwell::{
types::IntType,
values::{IntValue, PointerValue},
AddressSpace,
};
use itertools::Itertools;
use nac3parser::ast::{Expr, ExprKind};
use crate::{
codegen::{
irrt,
types::{
ndarray::{NDArrayType, NDIndexType},
structure::StructField,
utils::SliceType,
},
values::{ndarray::NDArrayValue, utils::RustSlice, ProxyValue},
CodeGenContext, CodeGenerator,
},
typecheck::typedef::Type,
};
/// An IRRT representation of an ndarray subscript index.
#[derive(Copy, Clone)]
pub struct NDIndexValue<'ctx> {
value: PointerValue<'ctx>,
llvm_usize: IntType<'ctx>,
name: Option<&'ctx str>,
}
impl<'ctx> NDIndexValue<'ctx> {
/// Checks whether `value` is an instance of `ndindex`, returning [Err] if `value` is not an
/// instance.
pub fn is_representable(
value: PointerValue<'ctx>,
llvm_usize: IntType<'ctx>,
) -> Result<(), String> {
<Self as ProxyValue<'ctx>>::Type::is_representable(value.get_type(), llvm_usize)
}
/// Creates an [`NDIndexValue`] from a [`PointerValue`].
#[must_use]
pub fn from_pointer_value(
ptr: PointerValue<'ctx>,
llvm_usize: IntType<'ctx>,
name: Option<&'ctx str>,
) -> Self {
debug_assert!(Self::is_representable(ptr, llvm_usize).is_ok());
Self { value: ptr, llvm_usize, name }
}
fn type_field(&self) -> StructField<'ctx, IntValue<'ctx>> {
self.get_type().get_fields().type_
}
pub fn load_type(&self, ctx: &CodeGenContext<'ctx, '_>) -> IntValue<'ctx> {
self.type_field().get(ctx, self.value, self.name)
}
pub fn store_type(&self, ctx: &CodeGenContext<'ctx, '_>, value: IntValue<'ctx>) {
self.type_field().set(ctx, self.value, value, self.name);
}
fn data_field(&self) -> StructField<'ctx, PointerValue<'ctx>> {
self.get_type().get_fields().data
}
pub fn load_data(&self, ctx: &CodeGenContext<'ctx, '_>) -> PointerValue<'ctx> {
self.data_field().get(ctx, self.value, self.name)
}
pub fn store_data(&self, ctx: &CodeGenContext<'ctx, '_>, value: PointerValue<'ctx>) {
self.data_field().set(ctx, self.value, value, self.name);
}
}
impl<'ctx> ProxyValue<'ctx> for NDIndexValue<'ctx> {
type Base = PointerValue<'ctx>;
type Type = NDIndexType<'ctx>;
fn get_type(&self) -> Self::Type {
Self::Type::from_type(self.value.get_type(), self.llvm_usize)
}
fn as_base_value(&self) -> Self::Base {
self.value
}
}
impl<'ctx> From<NDIndexValue<'ctx>> for PointerValue<'ctx> {
fn from(value: NDIndexValue<'ctx>) -> Self {
value.as_base_value()
}
}
impl<'ctx> NDArrayValue<'ctx> {
/// Get the expected `ndims` after indexing with `indices`.
#[must_use]
fn deduce_ndims_after_indexing_with(&self, indices: &[RustNDIndex<'ctx>]) -> Option<u64> {
let mut ndims = self.ndims?;
for index in indices {
match index {
RustNDIndex::SingleElement(_) => {
ndims -= 1; // Single elements decrements ndims
}
RustNDIndex::NewAxis => {
ndims += 1; // `np.newaxis` / `none` adds a new axis
}
RustNDIndex::Ellipsis | RustNDIndex::Slice(_) => {}
}
}
Some(ndims)
}
/// Index into the ndarray, and return a newly-allocated view on this ndarray.
///
/// This function behaves like NumPy's ndarray indexing, but if the indices index
/// into a single element, an unsized ndarray is returned.
#[must_use]
pub fn index<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
indices: &[RustNDIndex<'ctx>],
) -> Self {
assert!(self.ndims.is_some(), "NDArrayValue::index is only supported for instances with compile-time known ndims (self.ndims = Some(...))");
let dst_ndims = self.deduce_ndims_after_indexing_with(indices);
let dst_ndarray = NDArrayType::new(generator, ctx.ctx, self.dtype, dst_ndims)
.construct_uninitialized(generator, ctx, None);
let indices =
NDIndexType::new(generator, ctx.ctx).construct_ndindices(generator, ctx, indices);
irrt::ndarray::call_nac3_ndarray_index(generator, ctx, indices, *self, dst_ndarray);
dst_ndarray
}
}
/// A convenience enum representing a [`NDIndexValue`].
// TODO: Rename to CTConstNDIndex
#[derive(Debug, Clone)]
pub enum RustNDIndex<'ctx> {
SingleElement(IntValue<'ctx>),
Slice(RustSlice<'ctx>),
NewAxis,
Ellipsis,
}
impl<'ctx> RustNDIndex<'ctx> {
/// Generate LLVM code to transform an ndarray subscript expression to
/// its list of [`RustNDIndex`]
///
/// i.e.,
/// ```python
/// my_ndarray[::3, 1, :2:]
/// ^^^^^^^^^^^ Then these into a three `RustNDIndex`es
/// ```
pub fn from_subscript_expr<G: CodeGenerator>(
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
subscript: &Expr<Option<Type>>,
) -> Result<Vec<RustNDIndex<'ctx>>, String> {
// Annoying notes about `slice`
// - `my_array[5]`
// - slice is a `Constant`
// - `my_array[:5]`
// - slice is a `Slice`
// - `my_array[:]`
// - slice is a `Slice`, but lower upper step would all be `Option::None`
// - `my_array[:, :]`
// - slice is now a `Tuple` of two `Slice`-s
//
// In summary:
// - when there is a comma "," within [], `slice` will be a `Tuple` of the entries.
// - when there is not comma "," within [] (i.e., just a single entry), `slice` will be that entry itself.
//
// So we first "flatten" out the slice expression
let index_exprs = match &subscript.node {
ExprKind::Tuple { elts, .. } => elts.iter().collect_vec(),
_ => vec![subscript],
};
// Process all index expressions
let mut rust_ndindices: Vec<RustNDIndex> = Vec::with_capacity(index_exprs.len()); // Not using iterators here because `?` is used here.
for index_expr in index_exprs {
// NOTE: Currently nac3core's slices do not have an object representation,
// so the code/implementation looks awkward - we have to do pattern matching on the expression
let ndindex = if let ExprKind::Slice { lower, upper, step } = &index_expr.node {
// Handle slices
let slice = RustSlice::from_slice_expr(generator, ctx, lower, upper, step)?;
RustNDIndex::Slice(slice)
} else {
// Treat and handle everything else as a single element index.
let index = generator.gen_expr(ctx, index_expr)?.unwrap().to_basic_value_enum(
ctx,
generator,
ctx.primitives.int32, // Must be int32, this checks for illegal values
)?;
let index = index.into_int_value();
RustNDIndex::SingleElement(index)
};
rust_ndindices.push(ndindex);
}
Ok(rust_ndindices)
}
/// Get the value to set `NDIndex::type` for this variant.
#[must_use]
pub fn get_type_id(&self) -> u64 {
// Defined in IRRT, must be in sync
match self {
RustNDIndex::SingleElement(_) => 0,
RustNDIndex::Slice(_) => 1,
RustNDIndex::NewAxis => 2,
RustNDIndex::Ellipsis => 3,
}
}
/// Serialize this [`RustNDIndex`] by writing it into an LLVM [`NDIndexValue`].
pub fn write_to_ndindex<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
dst_ndindex: NDIndexValue<'ctx>,
) {
let llvm_pi8 = ctx.ctx.i8_type().ptr_type(AddressSpace::default());
// Set `dst_ndindex.type`
dst_ndindex.store_type(ctx, ctx.ctx.i8_type().const_int(self.get_type_id(), false));
// Set `dst_ndindex_ptr->data`
match self {
RustNDIndex::SingleElement(in_index) => {
let index_ptr = ctx.builder.build_alloca(ctx.ctx.i32_type(), "").unwrap();
ctx.builder.build_store(index_ptr, *in_index).unwrap();
dst_ndindex.store_data(
ctx,
ctx.builder.build_pointer_cast(index_ptr, llvm_pi8, "").unwrap(),
);
}
RustNDIndex::Slice(in_rust_slice) => {
let user_slice_ptr =
SliceType::new(ctx.ctx, ctx.ctx.i32_type(), generator.get_size_type(ctx.ctx))
.alloca(generator, ctx, None);
in_rust_slice.write_to_slice(ctx, user_slice_ptr);
dst_ndindex.store_data(
ctx,
ctx.builder.build_pointer_cast(user_slice_ptr.into(), llvm_pi8, "").unwrap(),
);
}
RustNDIndex::NewAxis | RustNDIndex::Ellipsis => {}
}
}
}

View File

@ -0,0 +1,933 @@
use inkwell::{
types::{AnyType, AnyTypeEnum, BasicType, BasicTypeEnum, IntType},
values::{BasicValueEnum, IntValue, PointerValue},
AddressSpace, IntPredicate,
};
use super::{
ArrayLikeIndexer, ArrayLikeValue, ProxyValue, TypedArrayLikeAccessor, TypedArrayLikeMutator,
UntypedArrayLikeAccessor, UntypedArrayLikeMutator,
};
use crate::codegen::{
irrt,
llvm_intrinsics::{call_int_umin, call_memcpy_generic_array},
stmt::gen_for_callback_incrementing,
type_aligned_alloca,
types::{ndarray::NDArrayType, structure::StructField},
CodeGenContext, CodeGenerator,
};
pub use contiguous::*;
pub use indexing::*;
pub use nditer::*;
pub use view::*;
mod contiguous;
mod indexing;
mod nditer;
mod view;
/// Proxy type for accessing an `NDArray` value in LLVM.
#[derive(Copy, Clone)]
pub struct NDArrayValue<'ctx> {
value: PointerValue<'ctx>,
dtype: BasicTypeEnum<'ctx>,
ndims: Option<u64>,
llvm_usize: IntType<'ctx>,
name: Option<&'ctx str>,
}
impl<'ctx> NDArrayValue<'ctx> {
/// Checks whether `value` is an instance of `NDArray`, returning [Err] if `value` is not an
/// instance.
pub fn is_representable(
value: PointerValue<'ctx>,
llvm_usize: IntType<'ctx>,
) -> Result<(), String> {
NDArrayType::is_representable(value.get_type(), llvm_usize)
}
/// Creates an [`NDArrayValue`] from a [`PointerValue`].
#[must_use]
pub fn from_pointer_value(
ptr: PointerValue<'ctx>,
dtype: BasicTypeEnum<'ctx>,
ndims: Option<u64>,
llvm_usize: IntType<'ctx>,
name: Option<&'ctx str>,
) -> Self {
debug_assert!(Self::is_representable(ptr, llvm_usize).is_ok());
NDArrayValue { value: ptr, dtype, ndims, llvm_usize, name }
}
fn ndims_field(&self, ctx: &CodeGenContext<'ctx, '_>) -> StructField<'ctx, IntValue<'ctx>> {
self.get_type().get_fields(ctx.ctx).ndims
}
/// Returns the pointer to the field storing the number of dimensions of this `NDArray`.
fn ptr_to_ndims(&self, ctx: &CodeGenContext<'ctx, '_>) -> PointerValue<'ctx> {
self.ndims_field(ctx).ptr_by_gep(ctx, self.value, self.name)
}
/// Stores the number of dimensions `ndims` into this instance.
pub fn store_ndims<G: CodeGenerator + ?Sized>(
&self,
ctx: &CodeGenContext<'ctx, '_>,
generator: &G,
ndims: IntValue<'ctx>,
) {
debug_assert_eq!(ndims.get_type(), generator.get_size_type(ctx.ctx));
let pndims = self.ptr_to_ndims(ctx);
ctx.builder.build_store(pndims, ndims).unwrap();
}
/// Returns the number of dimensions of this `NDArray` as a value.
pub fn load_ndims(&self, ctx: &CodeGenContext<'ctx, '_>) -> IntValue<'ctx> {
let pndims = self.ptr_to_ndims(ctx);
ctx.builder.build_load(pndims, "").map(BasicValueEnum::into_int_value).unwrap()
}
fn itemsize_field(&self, ctx: &CodeGenContext<'ctx, '_>) -> StructField<'ctx, IntValue<'ctx>> {
self.get_type().get_fields(ctx.ctx).itemsize
}
/// Stores the size of each element `itemsize` into this instance.
pub fn store_itemsize<G: CodeGenerator + ?Sized>(
&self,
ctx: &CodeGenContext<'ctx, '_>,
generator: &G,
itemsize: IntValue<'ctx>,
) {
debug_assert_eq!(itemsize.get_type(), generator.get_size_type(ctx.ctx));
self.itemsize_field(ctx).set(ctx, self.value, itemsize, self.name);
}
/// Returns the size of each element of this `NDArray` as a value.
pub fn load_itemsize(&self, ctx: &CodeGenContext<'ctx, '_>) -> IntValue<'ctx> {
self.itemsize_field(ctx).get(ctx, self.value, self.name)
}
fn shape_field(&self, ctx: &CodeGenContext<'ctx, '_>) -> StructField<'ctx, PointerValue<'ctx>> {
self.get_type().get_fields(ctx.ctx).shape
}
/// Returns the double-indirection pointer to the `shape` array, as if by calling
/// `getelementptr` on the field.
fn ptr_to_shape(&self, ctx: &CodeGenContext<'ctx, '_>) -> PointerValue<'ctx> {
self.shape_field(ctx).ptr_by_gep(ctx, self.value, self.name)
}
/// Stores the array of dimension sizes `dims` into this instance.
fn store_shape(&self, ctx: &CodeGenContext<'ctx, '_>, dims: PointerValue<'ctx>) {
self.shape_field(ctx).set(ctx, self.as_base_value(), dims, self.name);
}
/// Convenience method for creating a new array storing dimension sizes with the given `size`.
pub fn create_shape(
&self,
ctx: &CodeGenContext<'ctx, '_>,
llvm_usize: IntType<'ctx>,
size: IntValue<'ctx>,
) {
self.store_shape(ctx, ctx.builder.build_array_alloca(llvm_usize, size, "").unwrap());
}
/// Returns a proxy object to the field storing the size of each dimension of this `NDArray`.
#[must_use]
pub fn shape(&self) -> NDArrayShapeProxy<'ctx, '_> {
NDArrayShapeProxy(self)
}
fn strides_field(
&self,
ctx: &CodeGenContext<'ctx, '_>,
) -> StructField<'ctx, PointerValue<'ctx>> {
self.get_type().get_fields(ctx.ctx).strides
}
/// Returns the double-indirection pointer to the `strides` array, as if by calling
/// `getelementptr` on the field.
fn ptr_to_strides(&self, ctx: &CodeGenContext<'ctx, '_>) -> PointerValue<'ctx> {
self.strides_field(ctx).ptr_by_gep(ctx, self.value, self.name)
}
/// Stores the array of stride sizes `strides` into this instance.
fn store_strides(&self, ctx: &CodeGenContext<'ctx, '_>, strides: PointerValue<'ctx>) {
self.strides_field(ctx).set(ctx, self.as_base_value(), strides, self.name);
}
/// Convenience method for creating a new array storing the stride with the given `size`.
pub fn create_strides(
&self,
ctx: &CodeGenContext<'ctx, '_>,
llvm_usize: IntType<'ctx>,
size: IntValue<'ctx>,
) {
self.store_strides(ctx, ctx.builder.build_array_alloca(llvm_usize, size, "").unwrap());
}
/// Returns a proxy object to the field storing the stride of each dimension of this `NDArray`.
#[must_use]
pub fn strides(&self) -> NDArrayStridesProxy<'ctx, '_> {
NDArrayStridesProxy(self)
}
fn data_field(&self, ctx: &CodeGenContext<'ctx, '_>) -> StructField<'ctx, PointerValue<'ctx>> {
self.get_type().get_fields(ctx.ctx).data
}
/// Returns the double-indirection pointer to the `data` array, as if by calling `getelementptr`
/// on the field.
pub fn ptr_to_data(&self, ctx: &CodeGenContext<'ctx, '_>) -> PointerValue<'ctx> {
self.data_field(ctx).ptr_by_gep(ctx, self.value, self.name)
}
/// Stores the array of data elements `data` into this instance.
fn store_data(&self, ctx: &CodeGenContext<'ctx, '_>, data: PointerValue<'ctx>) {
let data = ctx
.builder
.build_bit_cast(data, ctx.ctx.i8_type().ptr_type(AddressSpace::default()), "")
.unwrap();
self.data_field(ctx).set(ctx, self.as_base_value(), data.into_pointer_value(), self.name);
}
/// Convenience method for creating a new array storing data elements with the given element
/// type `elem_ty` and `size`.
///
/// The data buffer will be allocated on the stack, and is considered to be owned by this ndarray instance.
///
/// # Safety
///
/// The caller must ensure that `shape` and `itemsize` of this ndarray instance is initialized.
pub unsafe fn create_data<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
) {
let nbytes = self.nbytes(generator, ctx);
let data = type_aligned_alloca(generator, ctx, self.dtype, nbytes, None);
self.store_data(ctx, data);
self.set_strides_contiguous(generator, ctx);
}
/// Returns a proxy object to the field storing the data of this `NDArray`.
#[must_use]
pub fn data(&self) -> NDArrayDataProxy<'ctx, '_> {
NDArrayDataProxy(self)
}
/// Copy shape dimensions from an array.
pub fn copy_shape_from_array<G: CodeGenerator + ?Sized>(
&self,
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
shape: PointerValue<'ctx>,
) {
let num_items = self.load_ndims(ctx);
call_memcpy_generic_array(
ctx,
self.shape().base_ptr(ctx, generator),
shape,
num_items,
ctx.ctx.bool_type().const_zero(),
);
}
/// Copy shape dimensions from an ndarray.
/// Panics if `ndims` mismatches.
pub fn copy_shape_from_ndarray<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
src_ndarray: NDArrayValue<'ctx>,
) {
if self.ndims.is_some() && src_ndarray.ndims.is_some() {
assert_eq!(self.ndims, src_ndarray.ndims);
} else {
let self_ndims = self.load_ndims(ctx);
let src_ndims = src_ndarray.load_ndims(ctx);
ctx.make_assert(
generator,
ctx.builder.build_int_compare(
IntPredicate::EQ,
self_ndims,
src_ndims,
""
).unwrap(),
"0:AssertionError",
"NDArrayValue::copy_shape_from_ndarray: Expected self.ndims ({0}) == src_ndarray.ndims ({1})",
[Some(self_ndims), Some(src_ndims), None],
ctx.current_loc
);
}
let src_shape = src_ndarray.shape().base_ptr(ctx, generator);
self.copy_shape_from_array(generator, ctx, src_shape);
}
/// Copy strides dimensions from an array.
pub fn copy_strides_from_array<G: CodeGenerator + ?Sized>(
&self,
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
strides: PointerValue<'ctx>,
) {
let num_items = self.load_ndims(ctx);
call_memcpy_generic_array(
ctx,
self.strides().base_ptr(ctx, generator),
strides,
num_items,
ctx.ctx.bool_type().const_zero(),
);
}
/// Copy strides dimensions from an ndarray.
/// Panics if `ndims` mismatches.
pub fn copy_strides_from_ndarray<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
src_ndarray: NDArrayValue<'ctx>,
) {
if self.ndims.is_some() && src_ndarray.ndims.is_some() {
assert_eq!(self.ndims, src_ndarray.ndims);
} else {
let self_ndims = self.load_ndims(ctx);
let src_ndims = src_ndarray.load_ndims(ctx);
ctx.make_assert(
generator,
ctx.builder.build_int_compare(
IntPredicate::EQ,
self_ndims,
src_ndims,
""
).unwrap(),
"0:AssertionError",
"NDArrayValue::copy_shape_from_ndarray: Expected self.ndims ({0}) == src_ndarray.ndims ({1})",
[Some(self_ndims), Some(src_ndims), None],
ctx.current_loc
);
}
let src_strides = src_ndarray.strides().base_ptr(ctx, generator);
self.copy_strides_from_array(generator, ctx, src_strides);
}
/// Get the `np.size()` of this ndarray.
pub fn size<G: CodeGenerator + ?Sized>(
&self,
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
) -> IntValue<'ctx> {
irrt::ndarray::call_nac3_ndarray_size(generator, ctx, *self)
}
/// Get the `ndarray.nbytes` of this ndarray.
pub fn nbytes<G: CodeGenerator + ?Sized>(
&self,
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
) -> IntValue<'ctx> {
irrt::ndarray::call_nac3_ndarray_nbytes(generator, ctx, *self)
}
/// Get the `len()` of this ndarray.
pub fn len<G: CodeGenerator + ?Sized>(
&self,
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
) -> IntValue<'ctx> {
irrt::ndarray::call_nac3_ndarray_len(generator, ctx, *self)
}
/// Check if this ndarray is C-contiguous.
///
/// See NumPy's `flags["C_CONTIGUOUS"]`: <https://numpy.org/doc/stable/reference/generated/numpy.ndarray.flags.html#numpy.ndarray.flags>
pub fn is_c_contiguous<G: CodeGenerator + ?Sized>(
&self,
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
) -> IntValue<'ctx> {
irrt::ndarray::call_nac3_ndarray_is_c_contiguous(generator, ctx, *self)
}
/// Call [`call_nac3_ndarray_set_strides_by_shape`] on this ndarray to update `strides`.
///
/// Update the ndarray's strides to make the ndarray contiguous.
pub fn set_strides_contiguous<G: CodeGenerator + ?Sized>(
&self,
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
) {
irrt::ndarray::call_nac3_ndarray_set_strides_by_shape(generator, ctx, *self);
}
#[must_use]
pub fn make_copy<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
) -> Self {
let clone = if self.ndims.is_some() {
self.get_type().construct_uninitialized(generator, ctx, None)
} else {
self.get_type().construct_dyn_ndims(generator, ctx, self.load_ndims(ctx), None)
};
let shape = self.shape();
clone.copy_shape_from_array(generator, ctx, shape.base_ptr(ctx, generator));
unsafe { clone.create_data(generator, ctx) };
clone.copy_data_from(generator, ctx, *self);
clone
}
/// Copy data from another ndarray.
///
/// This ndarray and `src` is that their `np.size()` should be the same. Their shapes
/// do not matter. The copying order is determined by how their flattened views look.
///
/// Panics if the `dtype`s of ndarrays are different.
pub fn copy_data_from<G: CodeGenerator + ?Sized>(
&self,
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
src: NDArrayValue<'ctx>,
) {
assert_eq!(self.dtype, src.dtype, "self and src dtype should match");
irrt::ndarray::call_nac3_ndarray_copy_data(generator, ctx, src, *self);
}
/// Returns true if this ndarray is unsized - `ndims == 0` and only contains a scalar.
#[must_use]
pub fn is_unsized(&self) -> Option<bool> {
self.ndims.map(|ndims| ndims == 0)
}
/// If this ndarray is unsized, return its sole value as an [`AnyObject`].
/// Otherwise, do nothing and return the ndarray itself.
// TODO: Rename to get_unsized_element
pub fn split_unsized<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
) -> ScalarOrNDArray<'ctx> {
let Some(is_unsized) = self.is_unsized() else { todo!() };
if is_unsized {
// NOTE: `np.size(self) == 0` here is never possible.
let zero = generator.get_size_type(ctx.ctx).const_zero();
let value = unsafe { self.data().get_unchecked(ctx, generator, &zero, None) };
ScalarOrNDArray::Scalar(value)
} else {
ScalarOrNDArray::NDArray(*self)
}
}
}
impl<'ctx> ProxyValue<'ctx> for NDArrayValue<'ctx> {
type Base = PointerValue<'ctx>;
type Type = NDArrayType<'ctx>;
fn get_type(&self) -> Self::Type {
NDArrayType::from_type(
self.as_base_value().get_type(),
self.dtype,
self.ndims,
self.llvm_usize,
)
}
fn as_base_value(&self) -> Self::Base {
self.value
}
}
impl<'ctx> From<NDArrayValue<'ctx>> for PointerValue<'ctx> {
fn from(value: NDArrayValue<'ctx>) -> Self {
value.as_base_value()
}
}
/// Proxy type for accessing the `shape` array of an `NDArray` instance in LLVM.
#[derive(Copy, Clone)]
pub struct NDArrayShapeProxy<'ctx, 'a>(&'a NDArrayValue<'ctx>);
impl<'ctx> ArrayLikeValue<'ctx> for NDArrayShapeProxy<'ctx, '_> {
fn element_type<G: CodeGenerator + ?Sized>(
&self,
ctx: &CodeGenContext<'ctx, '_>,
generator: &G,
) -> AnyTypeEnum<'ctx> {
self.0.shape().base_ptr(ctx, generator).get_type().get_element_type()
}
fn base_ptr<G: CodeGenerator + ?Sized>(
&self,
ctx: &CodeGenContext<'ctx, '_>,
_: &G,
) -> PointerValue<'ctx> {
self.0.shape_field(ctx).get(ctx, self.0.as_base_value(), self.0.name)
}
fn size<G: CodeGenerator + ?Sized>(
&self,
ctx: &CodeGenContext<'ctx, '_>,
_: &G,
) -> IntValue<'ctx> {
self.0.load_ndims(ctx)
}
}
impl<'ctx> ArrayLikeIndexer<'ctx, IntValue<'ctx>> for NDArrayShapeProxy<'ctx, '_> {
unsafe fn ptr_offset_unchecked<G: CodeGenerator + ?Sized>(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
idx: &IntValue<'ctx>,
name: Option<&str>,
) -> PointerValue<'ctx> {
let var_name = name.map(|v| format!("{v}.addr")).unwrap_or_default();
unsafe {
ctx.builder
.build_in_bounds_gep(self.base_ptr(ctx, generator), &[*idx], var_name.as_str())
.unwrap()
}
}
fn ptr_offset<G: CodeGenerator + ?Sized>(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
idx: &IntValue<'ctx>,
name: Option<&str>,
) -> PointerValue<'ctx> {
let size = self.size(ctx, generator);
let in_range = ctx.builder.build_int_compare(IntPredicate::ULT, *idx, size, "").unwrap();
ctx.make_assert(
generator,
in_range,
"0:IndexError",
"index {0} is out of bounds for axis 0 with size {1}",
[Some(*idx), Some(self.0.load_ndims(ctx)), None],
ctx.current_loc,
);
unsafe { self.ptr_offset_unchecked(ctx, generator, idx, name) }
}
}
impl<'ctx> UntypedArrayLikeAccessor<'ctx, IntValue<'ctx>> for NDArrayShapeProxy<'ctx, '_> {}
impl<'ctx> UntypedArrayLikeMutator<'ctx, IntValue<'ctx>> for NDArrayShapeProxy<'ctx, '_> {}
impl<'ctx> TypedArrayLikeAccessor<'ctx, IntValue<'ctx>> for NDArrayShapeProxy<'ctx, '_> {
fn downcast_to_type(
&self,
_: &mut CodeGenContext<'ctx, '_>,
value: BasicValueEnum<'ctx>,
) -> IntValue<'ctx> {
value.into_int_value()
}
}
impl<'ctx> TypedArrayLikeMutator<'ctx, IntValue<'ctx>> for NDArrayShapeProxy<'ctx, '_> {
fn upcast_from_type(
&self,
_: &mut CodeGenContext<'ctx, '_>,
value: IntValue<'ctx>,
) -> BasicValueEnum<'ctx> {
value.into()
}
}
/// Proxy type for accessing the `strides` array of an `NDArray` instance in LLVM.
#[derive(Copy, Clone)]
pub struct NDArrayStridesProxy<'ctx, 'a>(&'a NDArrayValue<'ctx>);
impl<'ctx> ArrayLikeValue<'ctx> for NDArrayStridesProxy<'ctx, '_> {
fn element_type<G: CodeGenerator + ?Sized>(
&self,
ctx: &CodeGenContext<'ctx, '_>,
generator: &G,
) -> AnyTypeEnum<'ctx> {
self.0.strides().base_ptr(ctx, generator).get_type().get_element_type()
}
fn base_ptr<G: CodeGenerator + ?Sized>(
&self,
ctx: &CodeGenContext<'ctx, '_>,
_: &G,
) -> PointerValue<'ctx> {
self.0.strides_field(ctx).get(ctx, self.0.as_base_value(), self.0.name)
}
fn size<G: CodeGenerator + ?Sized>(
&self,
ctx: &CodeGenContext<'ctx, '_>,
_: &G,
) -> IntValue<'ctx> {
self.0.load_ndims(ctx)
}
}
impl<'ctx> ArrayLikeIndexer<'ctx, IntValue<'ctx>> for NDArrayStridesProxy<'ctx, '_> {
unsafe fn ptr_offset_unchecked<G: CodeGenerator + ?Sized>(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
idx: &IntValue<'ctx>,
name: Option<&str>,
) -> PointerValue<'ctx> {
let var_name = name.map(|v| format!("{v}.addr")).unwrap_or_default();
unsafe {
ctx.builder
.build_in_bounds_gep(self.base_ptr(ctx, generator), &[*idx], var_name.as_str())
.unwrap()
}
}
fn ptr_offset<G: CodeGenerator + ?Sized>(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
idx: &IntValue<'ctx>,
name: Option<&str>,
) -> PointerValue<'ctx> {
let size = self.size(ctx, generator);
let in_range = ctx.builder.build_int_compare(IntPredicate::ULT, *idx, size, "").unwrap();
ctx.make_assert(
generator,
in_range,
"0:IndexError",
"index {0} is out of bounds for axis 0 with size {1}",
[Some(*idx), Some(self.0.load_ndims(ctx)), None],
ctx.current_loc,
);
unsafe { self.ptr_offset_unchecked(ctx, generator, idx, name) }
}
}
impl<'ctx> UntypedArrayLikeAccessor<'ctx, IntValue<'ctx>> for NDArrayStridesProxy<'ctx, '_> {}
impl<'ctx> UntypedArrayLikeMutator<'ctx, IntValue<'ctx>> for NDArrayStridesProxy<'ctx, '_> {}
impl<'ctx> TypedArrayLikeAccessor<'ctx, IntValue<'ctx>> for NDArrayStridesProxy<'ctx, '_> {
fn downcast_to_type(
&self,
_: &mut CodeGenContext<'ctx, '_>,
value: BasicValueEnum<'ctx>,
) -> IntValue<'ctx> {
value.into_int_value()
}
}
impl<'ctx> TypedArrayLikeMutator<'ctx, IntValue<'ctx>> for NDArrayStridesProxy<'ctx, '_> {
fn upcast_from_type(
&self,
_: &mut CodeGenContext<'ctx, '_>,
value: IntValue<'ctx>,
) -> BasicValueEnum<'ctx> {
value.into()
}
}
/// Proxy type for accessing the `data` array of an `NDArray` instance in LLVM.
#[derive(Copy, Clone)]
pub struct NDArrayDataProxy<'ctx, 'a>(&'a NDArrayValue<'ctx>);
impl<'ctx> ArrayLikeValue<'ctx> for NDArrayDataProxy<'ctx, '_> {
fn element_type<G: CodeGenerator + ?Sized>(
&self,
_: &CodeGenContext<'ctx, '_>,
_: &G,
) -> AnyTypeEnum<'ctx> {
self.0.dtype.as_any_type_enum()
}
fn base_ptr<G: CodeGenerator + ?Sized>(
&self,
ctx: &CodeGenContext<'ctx, '_>,
_: &G,
) -> PointerValue<'ctx> {
self.0.data_field(ctx).get(ctx, self.0.as_base_value(), self.0.name)
}
fn size<G: CodeGenerator + ?Sized>(
&self,
ctx: &CodeGenContext<'ctx, '_>,
generator: &G,
) -> IntValue<'ctx> {
irrt::ndarray::call_ndarray_calc_size(
generator,
ctx,
&self.as_slice_value(ctx, generator),
(None, None),
)
}
}
impl<'ctx> ArrayLikeIndexer<'ctx> for NDArrayDataProxy<'ctx, '_> {
unsafe fn ptr_offset_unchecked<G: CodeGenerator + ?Sized>(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
idx: &IntValue<'ctx>,
name: Option<&str>,
) -> PointerValue<'ctx> {
let sizeof_elem = ctx
.builder
.build_int_truncate_or_bit_cast(
self.element_type(ctx, generator).size_of().unwrap(),
idx.get_type(),
"",
)
.unwrap();
let idx = ctx.builder.build_int_mul(*idx, sizeof_elem, "").unwrap();
let ptr = unsafe {
ctx.builder
.build_in_bounds_gep(
self.base_ptr(ctx, generator),
&[idx],
name.unwrap_or_default(),
)
.unwrap()
};
// Current implementation is transparent - The returned pointer type is
// already cast into the expected type, allowing for immediately
// load/store.
ctx.builder
.build_pointer_cast(
ptr,
BasicTypeEnum::try_from(self.element_type(ctx, generator))
.unwrap()
.ptr_type(AddressSpace::default()),
"",
)
.unwrap()
}
fn ptr_offset<G: CodeGenerator + ?Sized>(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
idx: &IntValue<'ctx>,
name: Option<&str>,
) -> PointerValue<'ctx> {
let data_sz = self.size(ctx, generator);
let in_range = ctx.builder.build_int_compare(IntPredicate::ULT, *idx, data_sz, "").unwrap();
ctx.make_assert(
generator,
in_range,
"0:IndexError",
"index {0} is out of bounds with size {1}",
[Some(*idx), Some(self.0.load_ndims(ctx)), None],
ctx.current_loc,
);
let ptr = unsafe { self.ptr_offset_unchecked(ctx, generator, idx, name) };
// Current implementation is transparent - The returned pointer type is
// already cast into the expected type, allowing for immediately
// load/store.
ctx.builder
.build_pointer_cast(
ptr,
BasicTypeEnum::try_from(self.element_type(ctx, generator))
.unwrap()
.ptr_type(AddressSpace::default()),
"",
)
.unwrap()
}
}
impl<'ctx> UntypedArrayLikeAccessor<'ctx, IntValue<'ctx>> for NDArrayDataProxy<'ctx, '_> {}
impl<'ctx> UntypedArrayLikeMutator<'ctx, IntValue<'ctx>> for NDArrayDataProxy<'ctx, '_> {}
impl<'ctx, Index: UntypedArrayLikeAccessor<'ctx>> ArrayLikeIndexer<'ctx, Index>
for NDArrayDataProxy<'ctx, '_>
{
unsafe fn ptr_offset_unchecked<G: CodeGenerator + ?Sized>(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
indices: &Index,
name: Option<&str>,
) -> PointerValue<'ctx> {
let llvm_usize = generator.get_size_type(ctx.ctx);
let indices_elem_ty = indices
.ptr_offset(ctx, generator, &llvm_usize.const_zero(), None)
.get_type()
.get_element_type();
let Ok(indices_elem_ty) = IntType::try_from(indices_elem_ty) else {
panic!("Expected list[int32] but got {indices_elem_ty}")
};
assert_eq!(
indices_elem_ty.get_bit_width(),
32,
"Expected list[int32] but got list[int{}]",
indices_elem_ty.get_bit_width()
);
let index = irrt::ndarray::call_ndarray_flatten_index(generator, ctx, *self.0, indices);
let sizeof_elem = ctx
.builder
.build_int_truncate_or_bit_cast(
self.element_type(ctx, generator).size_of().unwrap(),
index.get_type(),
"",
)
.unwrap();
let index = ctx.builder.build_int_mul(index, sizeof_elem, "").unwrap();
let ptr = unsafe {
ctx.builder
.build_in_bounds_gep(
self.base_ptr(ctx, generator),
&[index],
name.unwrap_or_default(),
)
.unwrap()
};
// TODO: Current implementation is transparent
ctx.builder
.build_pointer_cast(
ptr,
BasicTypeEnum::try_from(self.element_type(ctx, generator))
.unwrap()
.ptr_type(AddressSpace::default()),
"",
)
.unwrap()
}
fn ptr_offset<G: CodeGenerator + ?Sized>(
&self,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut G,
indices: &Index,
name: Option<&str>,
) -> PointerValue<'ctx> {
let llvm_usize = generator.get_size_type(ctx.ctx);
let indices_size = indices.size(ctx, generator);
let nidx_leq_ndims = ctx
.builder
.build_int_compare(IntPredicate::SLE, indices_size, self.0.load_ndims(ctx), "")
.unwrap();
ctx.make_assert(
generator,
nidx_leq_ndims,
"0:IndexError",
"invalid index to scalar variable",
[None, None, None],
ctx.current_loc,
);
let indices_len = indices.size(ctx, generator);
let ndarray_len = self.0.load_ndims(ctx);
let len = call_int_umin(ctx, indices_len, ndarray_len, None);
gen_for_callback_incrementing(
generator,
ctx,
None,
llvm_usize.const_zero(),
(len, false),
|generator, ctx, _, i| {
let (dim_idx, dim_sz) = unsafe {
(
indices.get_unchecked(ctx, generator, &i, None).into_int_value(),
self.0.shape().get_typed_unchecked(ctx, generator, &i, None),
)
};
let dim_idx = ctx
.builder
.build_int_z_extend_or_bit_cast(dim_idx, dim_sz.get_type(), "")
.unwrap();
let dim_lt =
ctx.builder.build_int_compare(IntPredicate::SLT, dim_idx, dim_sz, "").unwrap();
ctx.make_assert(
generator,
dim_lt,
"0:IndexError",
"index {0} is out of bounds for axis 0 with size {1}",
[Some(dim_idx), Some(dim_sz), None],
ctx.current_loc,
);
Ok(())
},
llvm_usize.const_int(1, false),
)
.unwrap();
let ptr = unsafe { self.ptr_offset_unchecked(ctx, generator, indices, name) };
// TODO: Current implementation is transparent
ctx.builder
.build_pointer_cast(
ptr,
BasicTypeEnum::try_from(self.element_type(ctx, generator))
.unwrap()
.ptr_type(AddressSpace::default()),
"",
)
.unwrap()
}
}
impl<'ctx, Index: UntypedArrayLikeAccessor<'ctx>> UntypedArrayLikeAccessor<'ctx, Index>
for NDArrayDataProxy<'ctx, '_>
{
}
impl<'ctx, Index: UntypedArrayLikeAccessor<'ctx>> UntypedArrayLikeMutator<'ctx, Index>
for NDArrayDataProxy<'ctx, '_>
{
}
/// A version of [`call_nac3_ndarray_set_strides_by_shape`] in Rust.
///
/// This function is used generating strides for globally defined contiguous ndarrays.
#[must_use]
pub fn make_contiguous_strides(itemsize: u64, ndims: u64, shape: &[u64]) -> Vec<u64> {
let mut strides = Vec::with_capacity(ndims as usize);
let mut stride_product = 1u64;
for i in 0..ndims {
let axis = ndims - i - 1;
strides[axis as usize] = stride_product * itemsize;
stride_product *= shape[axis as usize];
}
strides
}
/// A convenience enum for implementing functions that acts on scalars or ndarrays or both.
#[derive(Clone, Copy)]
pub enum ScalarOrNDArray<'ctx> {
Scalar(BasicValueEnum<'ctx>),
NDArray(NDArrayValue<'ctx>),
}
impl<'ctx> ScalarOrNDArray<'ctx> {
/// Get the underlying [`BasicValueEnum<'ctx>`] of this [`ScalarOrNDArray`].
#[must_use]
pub fn to_basic_value_enum(self) -> BasicValueEnum<'ctx> {
match self {
ScalarOrNDArray::Scalar(scalar) => scalar,
ScalarOrNDArray::NDArray(ndarray) => ndarray.as_base_value().into(),
}
}
}

View File

@ -0,0 +1,176 @@
use inkwell::{
types::{BasicType, IntType},
values::{BasicValueEnum, IntValue, PointerValue},
AddressSpace,
};
use super::{NDArrayValue, ProxyValue, TypedArrayLikeAccessor, TypedArrayLikeMutator};
use crate::codegen::{
irrt,
stmt::{gen_for_callback, BreakContinueHooks},
types::{ndarray::NDIterType, structure::StructField},
values::{ArraySliceValue, TypedArrayLikeAdapter},
CodeGenContext, CodeGenerator,
};
#[derive(Copy, Clone)]
pub struct NDIterValue<'ctx> {
value: PointerValue<'ctx>,
parent: NDArrayValue<'ctx>,
indices: ArraySliceValue<'ctx>,
llvm_usize: IntType<'ctx>,
name: Option<&'ctx str>,
}
impl<'ctx> NDIterValue<'ctx> {
/// Checks whether `value` is an instance of `NDArray`, returning [Err] if `value` is not an
/// instance.
pub fn is_representable(
value: PointerValue<'ctx>,
llvm_usize: IntType<'ctx>,
) -> Result<(), String> {
<Self as ProxyValue>::Type::is_representable(value.get_type(), llvm_usize)
}
/// Creates an [`NDArrayValue`] from a [`PointerValue`].
#[must_use]
pub fn from_pointer_value(
ptr: PointerValue<'ctx>,
parent: NDArrayValue<'ctx>,
indices: ArraySliceValue<'ctx>,
llvm_usize: IntType<'ctx>,
name: Option<&'ctx str>,
) -> Self {
debug_assert!(Self::is_representable(ptr, llvm_usize).is_ok());
Self { value: ptr, parent, indices, llvm_usize, name }
}
/// Is the current iteration valid?
///
/// If true, then `element`, `indices` and `nth` contain details about the current element.
///
/// If `ndarray` is unsized, this returns true only for the first iteration.
/// If `ndarray` is 0-sized, this always returns false.
#[must_use]
pub fn has_element<G: CodeGenerator + ?Sized>(
&self,
generator: &G,
ctx: &CodeGenContext<'ctx, '_>,
) -> IntValue<'ctx> {
irrt::ndarray::call_nac3_nditer_has_element(generator, ctx, *self)
}
/// Go to the next element. If `has_element()` is false, then this has undefined behavior.
///
/// If `ndarray` is unsized, this can only be called once.
/// If `ndarray` is 0-sized, this can never be called.
pub fn next<G: CodeGenerator + ?Sized>(&self, generator: &G, ctx: &CodeGenContext<'ctx, '_>) {
irrt::ndarray::call_nac3_nditer_next(generator, ctx, *self);
}
fn element(&self, ctx: &CodeGenContext<'ctx, '_>) -> StructField<'ctx, PointerValue<'ctx>> {
self.get_type().get_fields(ctx.ctx).element
}
/// Get pointer to the current element.
#[must_use]
pub fn get_pointer(&self, ctx: &CodeGenContext<'ctx, '_>) -> PointerValue<'ctx> {
let elem_ty = self.parent.dtype;
let p = self.element(ctx).get(ctx, self.as_base_value(), None);
ctx.builder
.build_pointer_cast(p, elem_ty.ptr_type(AddressSpace::default()), "element")
.unwrap()
}
/// Get the value of the current element.
#[must_use]
pub fn get_scalar(&self, ctx: &CodeGenContext<'ctx, '_>) -> BasicValueEnum<'ctx> {
let p = self.get_pointer(ctx);
ctx.builder.build_load(p, "value").unwrap()
}
fn nth(&self, ctx: &CodeGenContext<'ctx, '_>) -> StructField<'ctx, IntValue<'ctx>> {
self.get_type().get_fields(ctx.ctx).nth
}
/// Get the index of the current element if this ndarray were a flat ndarray.
#[must_use]
pub fn get_index(&self, ctx: &CodeGenContext<'ctx, '_>) -> IntValue<'ctx> {
self.nth(ctx).get(ctx, self.as_base_value(), None)
}
/// Get the indices of the current element.
#[must_use]
pub fn get_indices(
&'ctx self,
) -> impl TypedArrayLikeAccessor<'ctx, IntValue<'ctx>> + TypedArrayLikeMutator<'ctx, IntValue<'ctx>>
{
TypedArrayLikeAdapter::from(
self.indices,
Box::new(|ctx, val| {
ctx.builder
.build_int_z_extend_or_bit_cast(val.into_int_value(), self.llvm_usize, "")
.unwrap()
}),
Box::new(|_, val| val.into()),
)
}
}
impl<'ctx> ProxyValue<'ctx> for NDIterValue<'ctx> {
type Base = PointerValue<'ctx>;
type Type = NDIterType<'ctx>;
fn get_type(&self) -> Self::Type {
NDIterType::from_type(self.as_base_value().get_type(), self.llvm_usize)
}
fn as_base_value(&self) -> Self::Base {
self.value
}
}
impl<'ctx> From<NDIterValue<'ctx>> for PointerValue<'ctx> {
fn from(value: NDIterValue<'ctx>) -> Self {
value.as_base_value()
}
}
impl<'ctx> NDArrayValue<'ctx> {
/// Iterate through every element in the ndarray.
///
/// `body` has access to [`BreakContinueHooks`] to short-circuit and [`NDIterValue`] to
/// get properties of the current iteration (e.g., the current element, indices, etc.)
pub fn foreach<'a, G, F>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, 'a>,
body: F,
) -> Result<(), String>
where
G: CodeGenerator + ?Sized,
F: FnOnce(
&mut G,
&mut CodeGenContext<'ctx, 'a>,
BreakContinueHooks<'ctx>,
NDIterValue<'ctx>,
) -> Result<(), String>,
{
gen_for_callback(
generator,
ctx,
Some("ndarray_foreach"),
|generator, ctx| {
Ok(NDIterType::new(generator, ctx.ctx).construct(generator, ctx, *self))
},
|generator, ctx, nditer| Ok(nditer.has_element(generator, ctx)),
|generator, ctx, hooks, nditer| body(generator, ctx, hooks, nditer),
|generator, ctx, nditer| {
nditer.next(generator, ctx);
Ok(())
},
)
}
}

View File

@ -0,0 +1,36 @@
use std::iter::{once, repeat_n};
use itertools::Itertools;
use crate::codegen::{
values::ndarray::{NDArrayValue, RustNDIndex},
CodeGenContext, CodeGenerator,
};
impl<'ctx> NDArrayValue<'ctx> {
/// Make sure the ndarray is at least `ndmin`-dimensional.
///
/// If this ndarray's `ndims` is less than `ndmin`, a view is created on this with 1s prepended
/// to the shape. Otherwise, this function does nothing and return this ndarray.
#[must_use]
pub fn atleast_nd<G: CodeGenerator + ?Sized>(
&self,
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
ndmin: u64,
) -> Self {
assert!(self.ndims.is_some(), "NDArrayValue::atleast_nd is only supported for instances with compile-time known ndims (self.ndims = Some(...))");
let ndims = self.ndims.unwrap();
if ndims < ndmin {
// Extend the dimensions with np.newaxis.
let indices = repeat_n(RustNDIndex::NewAxis, (ndmin - ndims) as usize)
.chain(once(RustNDIndex::Ellipsis))
.collect_vec();
self.index(generator, ctx, &indices)
} else {
*self
}
}
}

View File

@ -0,0 +1,153 @@
use inkwell::values::{BasicValueEnum, IntValue, PointerValue};
use super::ProxyValue;
use crate::codegen::{types::RangeType, CodeGenContext};
/// Proxy type for accessing a `range` value in LLVM.
#[derive(Copy, Clone)]
pub struct RangeValue<'ctx> {
value: PointerValue<'ctx>,
name: Option<&'ctx str>,
}
impl<'ctx> RangeValue<'ctx> {
/// Checks whether `value` is an instance of `range`, returning [Err] if `value` is not an instance.
pub fn is_representable(value: PointerValue<'ctx>) -> Result<(), String> {
RangeType::is_representable(value.get_type())
}
/// Creates an [`RangeValue`] from a [`PointerValue`].
#[must_use]
pub fn from_pointer_value(ptr: PointerValue<'ctx>, name: Option<&'ctx str>) -> Self {
debug_assert!(Self::is_representable(ptr).is_ok());
RangeValue { value: ptr, name }
}
fn ptr_to_start(&self, ctx: &CodeGenContext<'ctx, '_>) -> PointerValue<'ctx> {
let llvm_i32 = ctx.ctx.i32_type();
let var_name = self.name.map(|v| format!("{v}.start.addr")).unwrap_or_default();
unsafe {
ctx.builder
.build_in_bounds_gep(
self.as_base_value(),
&[llvm_i32.const_zero(), llvm_i32.const_int(0, false)],
var_name.as_str(),
)
.unwrap()
}
}
fn ptr_to_end(&self, ctx: &CodeGenContext<'ctx, '_>) -> PointerValue<'ctx> {
let llvm_i32 = ctx.ctx.i32_type();
let var_name = self.name.map(|v| format!("{v}.end.addr")).unwrap_or_default();
unsafe {
ctx.builder
.build_in_bounds_gep(
self.as_base_value(),
&[llvm_i32.const_zero(), llvm_i32.const_int(1, false)],
var_name.as_str(),
)
.unwrap()
}
}
fn ptr_to_step(&self, ctx: &CodeGenContext<'ctx, '_>) -> PointerValue<'ctx> {
let llvm_i32 = ctx.ctx.i32_type();
let var_name = self.name.map(|v| format!("{v}.step.addr")).unwrap_or_default();
unsafe {
ctx.builder
.build_in_bounds_gep(
self.as_base_value(),
&[llvm_i32.const_zero(), llvm_i32.const_int(2, false)],
var_name.as_str(),
)
.unwrap()
}
}
/// Stores the `start` value into this instance.
pub fn store_start(&self, ctx: &CodeGenContext<'ctx, '_>, start: IntValue<'ctx>) {
debug_assert_eq!(start.get_type().get_bit_width(), 32);
let pstart = self.ptr_to_start(ctx);
ctx.builder.build_store(pstart, start).unwrap();
}
/// Returns the `start` value of this `range`.
pub fn load_start(&self, ctx: &CodeGenContext<'ctx, '_>, name: Option<&str>) -> IntValue<'ctx> {
let pstart = self.ptr_to_start(ctx);
let var_name = name
.map(ToString::to_string)
.or_else(|| self.name.map(|v| format!("{v}.start")))
.unwrap_or_default();
ctx.builder
.build_load(pstart, var_name.as_str())
.map(BasicValueEnum::into_int_value)
.unwrap()
}
/// Stores the `end` value into this instance.
pub fn store_end(&self, ctx: &CodeGenContext<'ctx, '_>, end: IntValue<'ctx>) {
debug_assert_eq!(end.get_type().get_bit_width(), 32);
let pend = self.ptr_to_end(ctx);
ctx.builder.build_store(pend, end).unwrap();
}
/// Returns the `end` value of this `range`.
pub fn load_end(&self, ctx: &CodeGenContext<'ctx, '_>, name: Option<&str>) -> IntValue<'ctx> {
let pend = self.ptr_to_end(ctx);
let var_name = name
.map(ToString::to_string)
.or_else(|| self.name.map(|v| format!("{v}.end")))
.unwrap_or_default();
ctx.builder.build_load(pend, var_name.as_str()).map(BasicValueEnum::into_int_value).unwrap()
}
/// Stores the `step` value into this instance.
pub fn store_step(&self, ctx: &CodeGenContext<'ctx, '_>, step: IntValue<'ctx>) {
debug_assert_eq!(step.get_type().get_bit_width(), 32);
let pstep = self.ptr_to_step(ctx);
ctx.builder.build_store(pstep, step).unwrap();
}
/// Returns the `step` value of this `range`.
pub fn load_step(&self, ctx: &CodeGenContext<'ctx, '_>, name: Option<&str>) -> IntValue<'ctx> {
let pstep = self.ptr_to_step(ctx);
let var_name = name
.map(ToString::to_string)
.or_else(|| self.name.map(|v| format!("{v}.step")))
.unwrap_or_default();
ctx.builder
.build_load(pstep, var_name.as_str())
.map(BasicValueEnum::into_int_value)
.unwrap()
}
}
impl<'ctx> ProxyValue<'ctx> for RangeValue<'ctx> {
type Base = PointerValue<'ctx>;
type Type = RangeType<'ctx>;
fn get_type(&self) -> Self::Type {
RangeType::from_type(self.value.get_type())
}
fn as_base_value(&self) -> Self::Base {
self.value
}
}
impl<'ctx> From<RangeValue<'ctx>> for PointerValue<'ctx> {
fn from(value: RangeValue<'ctx>) -> Self {
value.as_base_value()
}
}

View File

@ -0,0 +1,3 @@
pub use slice::*;
mod slice;

View File

@ -0,0 +1,231 @@
use inkwell::{
types::IntType,
values::{IntValue, PointerValue},
};
use nac3parser::ast::Expr;
use crate::{
codegen::{
types::{structure::StructField, utils::SliceType},
values::ProxyValue,
CodeGenContext, CodeGenerator,
},
typecheck::typedef::Type,
};
/// An IRRT representation of an (unresolved) slice.
#[derive(Copy, Clone)]
pub struct SliceValue<'ctx> {
value: PointerValue<'ctx>,
int_ty: IntType<'ctx>,
llvm_usize: IntType<'ctx>,
name: Option<&'ctx str>,
}
impl<'ctx> SliceValue<'ctx> {
/// Checks whether `value` is an instance of `ContiguousNDArray`, returning [Err] if `value` is
/// not an instance.
pub fn is_representable(
value: PointerValue<'ctx>,
llvm_usize: IntType<'ctx>,
) -> Result<(), String> {
<Self as ProxyValue<'ctx>>::Type::is_representable(value.get_type(), llvm_usize)
}
/// Creates an [`SliceValue`] from a [`PointerValue`].
#[must_use]
pub fn from_pointer_value(
ptr: PointerValue<'ctx>,
int_ty: IntType<'ctx>,
llvm_usize: IntType<'ctx>,
name: Option<&'ctx str>,
) -> Self {
debug_assert!(Self::is_representable(ptr, llvm_usize).is_ok());
Self { value: ptr, int_ty, llvm_usize, name }
}
fn start_defined_field(&self) -> StructField<'ctx, IntValue<'ctx>> {
self.get_type().get_fields().start_defined
}
pub fn load_start_defined(&self, ctx: &CodeGenContext<'ctx, '_>) -> IntValue<'ctx> {
self.start_defined_field().get(ctx, self.value, self.name)
}
fn start_field(&self) -> StructField<'ctx, IntValue<'ctx>> {
self.get_type().get_fields().start
}
pub fn load_start(&self, ctx: &CodeGenContext<'ctx, '_>) -> IntValue<'ctx> {
self.start_field().get(ctx, self.value, self.name)
}
pub fn store_start(&self, ctx: &CodeGenContext<'ctx, '_>, value: Option<IntValue<'ctx>>) {
match value {
Some(start) => {
self.start_defined_field().set(
ctx,
self.value,
ctx.ctx.bool_type().const_all_ones(),
self.name,
);
self.start_field().set(ctx, self.value, start, self.name);
}
None => self.start_defined_field().set(
ctx,
self.value,
ctx.ctx.bool_type().const_zero(),
self.name,
),
}
}
fn stop_defined_field(&self) -> StructField<'ctx, IntValue<'ctx>> {
self.get_type().get_fields().stop_defined
}
pub fn load_stop_defined(&self, ctx: &CodeGenContext<'ctx, '_>) -> IntValue<'ctx> {
self.stop_defined_field().get(ctx, self.value, self.name)
}
fn stop_field(&self) -> StructField<'ctx, IntValue<'ctx>> {
self.get_type().get_fields().stop
}
pub fn load_stop(&self, ctx: &CodeGenContext<'ctx, '_>) -> IntValue<'ctx> {
self.stop_field().get(ctx, self.value, self.name)
}
pub fn store_stop(&self, ctx: &CodeGenContext<'ctx, '_>, value: Option<IntValue<'ctx>>) {
match value {
Some(stop) => {
self.stop_defined_field().set(
ctx,
self.value,
ctx.ctx.bool_type().const_all_ones(),
self.name,
);
self.stop_field().set(ctx, self.value, stop, self.name);
}
None => self.stop_defined_field().set(
ctx,
self.value,
ctx.ctx.bool_type().const_zero(),
self.name,
),
}
}
fn step_defined_field(&self) -> StructField<'ctx, IntValue<'ctx>> {
self.get_type().get_fields().step_defined
}
pub fn load_step_defined(&self, ctx: &CodeGenContext<'ctx, '_>) -> IntValue<'ctx> {
self.step_defined_field().get(ctx, self.value, self.name)
}
fn step_field(&self) -> StructField<'ctx, IntValue<'ctx>> {
self.get_type().get_fields().step
}
pub fn load_step(&self, ctx: &CodeGenContext<'ctx, '_>) -> IntValue<'ctx> {
self.step_field().get(ctx, self.value, self.name)
}
pub fn store_step(&self, ctx: &CodeGenContext<'ctx, '_>, value: Option<IntValue<'ctx>>) {
match value {
Some(step) => {
self.step_defined_field().set(
ctx,
self.value,
ctx.ctx.bool_type().const_all_ones(),
self.name,
);
self.step_field().set(ctx, self.value, step, self.name);
}
None => self.step_defined_field().set(
ctx,
self.value,
ctx.ctx.bool_type().const_zero(),
self.name,
),
}
}
}
impl<'ctx> ProxyValue<'ctx> for SliceValue<'ctx> {
type Base = PointerValue<'ctx>;
type Type = SliceType<'ctx>;
fn get_type(&self) -> Self::Type {
Self::Type::from_type(self.value.get_type(), self.int_ty, self.llvm_usize)
}
fn as_base_value(&self) -> Self::Base {
self.value
}
}
impl<'ctx> From<SliceValue<'ctx>> for PointerValue<'ctx> {
fn from(value: SliceValue<'ctx>) -> Self {
value.as_base_value()
}
}
/// A slice represented in compile-time by `start`, `stop` and `step`, all held as LLVM values.
// TODO: Rename this to CTConstSlice
#[derive(Debug, Copy, Clone)]
pub struct RustSlice<'ctx> {
int_ty: IntType<'ctx>,
start: Option<IntValue<'ctx>>,
stop: Option<IntValue<'ctx>>,
step: Option<IntValue<'ctx>>,
}
impl<'ctx> RustSlice<'ctx> {
/// Generate LLVM IR for an [`ExprKind::Slice`] and convert it into a [`RustSlice`].
#[allow(clippy::type_complexity)]
pub fn from_slice_expr<G: CodeGenerator>(
generator: &mut G,
ctx: &mut CodeGenContext<'ctx, '_>,
lower: &Option<Box<Expr<Option<Type>>>>,
upper: &Option<Box<Expr<Option<Type>>>>,
step: &Option<Box<Expr<Option<Type>>>>,
) -> Result<RustSlice<'ctx>, String> {
let mut value_mapper = |value_expr: &Option<Box<Expr<Option<Type>>>>| -> Result<_, String> {
Ok(match value_expr {
None => None,
Some(value_expr) => {
let value_expr = generator
.gen_expr(ctx, value_expr)?
.map(|value| {
value.to_basic_value_enum(ctx, generator, ctx.primitives.int32)
})
.unwrap()?;
Some(value_expr.into_int_value())
}
})
};
let start = value_mapper(lower)?;
let stop = value_mapper(upper)?;
let step = value_mapper(step)?;
Ok(RustSlice { int_ty: ctx.ctx.i32_type(), start, stop, step })
}
/// Write the contents to an LLVM [`SliceValue`].
pub fn write_to_slice(&self, ctx: &CodeGenContext<'ctx, '_>, dst_slice_ptr: SliceValue<'ctx>) {
assert_eq!(self.int_ty, dst_slice_ptr.int_ty);
dst_slice_ptr.store_start(ctx, self.start);
dst_slice_ptr.store_stop(ctx, self.stop);
dst_slice_ptr.store_step(ctx, self.step);
}
}

View File

@ -1,10 +1,4 @@
#![deny(
future_incompatible,
let_underscore,
nonstandard_style,
rust_2024_compatibility,
clippy::all
)]
#![deny(future_incompatible, let_underscore, nonstandard_style, clippy::all)]
#![warn(clippy::pedantic)]
#![allow(
dead_code,
@ -19,7 +13,13 @@
clippy::wildcard_imports
)]
// users of nac3core need to use the same version of these dependencies, so expose them as nac3core::*
pub use inkwell;
pub use nac3parser;
pub mod codegen;
pub mod symbol_resolver;
pub mod toplevel;
pub mod typecheck;
extern crate self as nac3core;

View File

@ -1,7 +1,15 @@
use std::fmt::Debug;
use std::rc::Rc;
use std::sync::Arc;
use std::{collections::HashMap, collections::HashSet, fmt::Display};
use std::{
collections::{HashMap, HashSet},
fmt::{Debug, Display},
rc::Rc,
sync::Arc,
};
use inkwell::values::{BasicValueEnum, FloatValue, IntValue, PointerValue, StructValue};
use itertools::{chain, izip, Itertools};
use parking_lot::RwLock;
use nac3parser::ast::{Constant, Expr, Location, StrRef};
use crate::{
codegen::{CodeGenContext, CodeGenerator},
@ -11,10 +19,6 @@ use crate::{
typedef::{Type, TypeEnum, Unifier, VarMap},
},
};
use inkwell::values::{BasicValueEnum, FloatValue, IntValue, PointerValue, StructValue};
use itertools::{chain, izip, Itertools};
use nac3parser::ast::{Constant, Expr, Location, StrRef};
use parking_lot::RwLock;
#[derive(Clone, PartialEq, Debug)]
pub enum SymbolValue {
@ -365,6 +369,7 @@ pub trait SymbolResolver {
&self,
str: StrRef,
ctx: &mut CodeGenContext<'ctx, '_>,
generator: &mut dyn CodeGenerator,
) -> Option<ValueEnum<'ctx>>;
fn get_default_param_value(&self, expr: &Expr) -> Option<SymbolValue>;

View File

@ -1,6 +1,5 @@
use std::iter::once;
use helper::{debug_assert_prim_is_allowed, make_exception_fields, PrimDefDetails};
use indexmap::IndexMap;
use inkwell::{
attributes::{Attribute, AttributeLoc},
@ -11,20 +10,22 @@ use inkwell::{
use itertools::Either;
use strum::IntoEnumIterator;
use super::{
helper::{debug_assert_prim_is_allowed, make_exception_fields, PrimDef, PrimDefDetails},
numpy::make_ndarray_ty,
*,
};
use crate::{
codegen::{
builtin_fns,
classes::{ProxyValue, RangeValue},
numpy::*,
stmt::exn_constructor,
values::{ProxyValue, RangeValue},
},
symbol_resolver::SymbolValue,
toplevel::{helper::PrimDef, numpy::make_ndarray_ty},
typecheck::typedef::{into_var_map, iter_type_vars, TypeVar, VarMap},
};
use super::*;
type BuiltinInfo = Vec<(Arc<RwLock<TopLevelDef>>, Option<Stmt>)>;
pub fn get_exn_constructor(
@ -709,7 +710,7 @@ impl<'a> BuiltinBuilder<'a> {
let (zelf_ty, zelf) = obj.unwrap();
let zelf =
zelf.to_basic_value_enum(ctx, generator, zelf_ty)?.into_pointer_value();
let zelf = RangeValue::from_ptr_val(zelf, Some("range"));
let zelf = RangeValue::from_pointer_value(zelf, Some("range"));
let mut start = None;
let mut stop = None;

File diff suppressed because it is too large Load Diff

View File

@ -1,14 +1,15 @@
use std::convert::TryInto;
use crate::symbol_resolver::SymbolValue;
use crate::toplevel::numpy::unpack_ndarray_var_tys;
use crate::typecheck::typedef::{into_var_map, iter_type_vars, Mapping, TypeVarId, VarMap};
use ast::ExprKind;
use nac3parser::ast::{Constant, Location};
use strum::IntoEnumIterator;
use strum_macros::EnumIter;
use super::*;
use nac3parser::ast::{Constant, ExprKind, Location};
use super::{numpy::unpack_ndarray_var_tys, *};
use crate::{
symbol_resolver::SymbolValue,
typecheck::typedef::{into_var_map, iter_type_vars, Mapping, TypeVarId, VarMap},
};
/// All primitive types and functions in nac3core.
#[derive(Clone, Copy, Debug, EnumIter, PartialEq, Eq)]
@ -388,6 +389,9 @@ impl TopLevelDef {
r
}
),
TopLevelDef::Variable { name, ty, .. } => {
format!("Variable {{ name: {name:?}, ty: {:?} }}", unifier.stringify(*ty),)
}
}
}
}
@ -589,6 +593,18 @@ impl TopLevelComposer {
}
}
#[must_use]
pub fn make_top_level_variable_def(
name: String,
simple_name: StrRef,
ty: Type,
ty_decl: Option<Expr>,
resolver: Option<Arc<dyn SymbolResolver + Send + Sync>>,
loc: Option<Location>,
) -> TopLevelDef {
TopLevelDef::Variable { name, simple_name, ty, ty_decl, resolver, loc }
}
#[must_use]
pub fn make_class_method_name(mut class_name: String, method_name: &str) -> String {
class_name.push('.');
@ -608,64 +624,6 @@ impl TopLevelComposer {
Err(HashSet::from([format!("no method {method_name} in the current class")]))
}
/// get all base class def id of a class, excluding itself. \
/// this function should called only after the direct parent is set
/// and before all the ancestors are set
/// and when we allow single inheritance \
/// the order of the returned list is from the child to the deepest ancestor
pub fn get_all_ancestors_helper(
child: &TypeAnnotation,
temp_def_list: &[Arc<RwLock<TopLevelDef>>],
) -> Result<Vec<TypeAnnotation>, HashSet<String>> {
let mut result: Vec<TypeAnnotation> = Vec::new();
let mut parent = Self::get_parent(child, temp_def_list);
while let Some(p) = parent {
parent = Self::get_parent(&p, temp_def_list);
let p_id = if let TypeAnnotation::CustomClass { id, .. } = &p {
*id
} else {
unreachable!("must be class kind annotation")
};
// check cycle
let no_cycle = result.iter().all(|x| {
let TypeAnnotation::CustomClass { id, .. } = x else {
unreachable!("must be class kind annotation")
};
id.0 != p_id.0
});
if no_cycle {
result.push(p);
} else {
return Err(HashSet::from(["cyclic inheritance detected".into()]));
}
}
Ok(result)
}
/// should only be called when finding all ancestors, so panic when wrong
fn get_parent(
child: &TypeAnnotation,
temp_def_list: &[Arc<RwLock<TopLevelDef>>],
) -> Option<TypeAnnotation> {
let child_id = if let TypeAnnotation::CustomClass { id, .. } = child {
*id
} else {
unreachable!("should be class type annotation")
};
let child_def = temp_def_list.get(child_id.0).unwrap();
let child_def = child_def.read();
let TopLevelDef::Class { ancestors, .. } = &*child_def else {
unreachable!("child must be top level class def")
};
if ancestors.is_empty() {
None
} else {
Some(ancestors[0].clone())
}
}
/// get the `var_id` of a given `TVar` type
pub fn get_var_id(var_ty: Type, unifier: &mut Unifier) -> Result<TypeVarId, HashSet<String>> {
if let TypeEnum::TVar { id, .. } = unifier.get_ty(var_ty).as_ref() {
@ -975,6 +933,139 @@ impl TopLevelComposer {
))
}
}
/// Parses the class type variables and direct parents
/// we only allow single inheritance
pub fn analyze_class_bases(
class_def: &Arc<RwLock<TopLevelDef>>,
class_ast: &Option<Stmt>,
temp_def_list: &[Arc<RwLock<TopLevelDef>>],
unifier: &mut Unifier,
primitives_store: &PrimitiveStore,
) -> Result<(), HashSet<String>> {
let mut class_def = class_def.write();
let (class_def_id, class_ancestors, class_bases_ast, class_type_vars, class_resolver) = {
let TopLevelDef::Class { object_id, ancestors, type_vars, resolver, .. } =
&mut *class_def
else {
unreachable!()
};
let Some(ast::Located { node: ast::StmtKind::ClassDef { bases, .. }, .. }) = class_ast
else {
unreachable!()
};
(object_id, ancestors, bases, type_vars, resolver.as_ref().unwrap().as_ref())
};
let mut is_generic = false;
let mut has_base = false;
// Check class bases for typevars
for b in class_bases_ast {
match &b.node {
// analyze typevars bounded to the class,
// only support things like `class A(Generic[T, V])`,
// things like `class A(Generic[T, V, ImportedModule.T])` is not supported
// i.e. only simple names are allowed in the subscript
// should update the TopLevelDef::Class.typevars and the TypeEnum::TObj.params
ast::ExprKind::Subscript { value, slice, .. } if matches!(&value.node, ast::ExprKind::Name { id, .. } if id == &"Generic".into()) =>
{
if is_generic {
return Err(HashSet::from([format!(
"only single Generic[...] is allowed (at {})",
b.location
)]));
}
is_generic = true;
let type_var_list: Vec<&ast::Expr<()>>;
// if `class A(Generic[T, V, G])`
if let ast::ExprKind::Tuple { elts, .. } = &slice.node {
type_var_list = elts.iter().collect_vec();
// `class A(Generic[T])`
} else {
type_var_list = vec![&**slice];
}
let type_vars = type_var_list
.into_iter()
.map(|e| {
class_resolver.parse_type_annotation(
temp_def_list,
unifier,
primitives_store,
e,
)
})
.collect::<Result<Vec<_>, _>>()?;
class_type_vars.extend(type_vars);
}
ast::ExprKind::Name { .. } | ast::ExprKind::Subscript { .. } => {
if has_base {
return Err(HashSet::from([format!("a class definition can only have at most one base class declaration and one generic declaration (at {})", b.location )]));
}
has_base = true;
// the function parse_ast_to make sure that no type var occurred in
// bast_ty if it is a CustomClassKind
let base_ty = parse_ast_to_type_annotation_kinds(
class_resolver,
temp_def_list,
unifier,
primitives_store,
b,
vec![(*class_def_id, class_type_vars.clone())]
.into_iter()
.collect::<HashMap<_, _>>(),
)?;
if let TypeAnnotation::CustomClass { .. } = &base_ty {
class_ancestors.push(base_ty);
} else {
return Err(HashSet::from([format!(
"class base declaration can only be custom class (at {})",
b.location
)]));
}
}
_ => {
return Err(HashSet::from([format!(
"unsupported statement in class defintion (at {})",
b.location
)]));
}
}
}
Ok(())
}
/// gets all ancestors of a class
pub fn analyze_class_ancestors(
class_def: &Arc<RwLock<TopLevelDef>>,
temp_def_list: &[Arc<RwLock<TopLevelDef>>],
) {
// Check if class has a direct parent
let mut class_def = class_def.write();
let TopLevelDef::Class { ancestors, type_vars, object_id, .. } = &mut *class_def else {
unreachable!()
};
let mut anc_set = HashMap::new();
if let Some(ancestor) = ancestors.first() {
let TypeAnnotation::CustomClass { id, .. } = ancestor else { unreachable!() };
let TopLevelDef::Class { ancestors: parent_ancestors, .. } =
&*temp_def_list[id.0].read()
else {
unreachable!()
};
for anc in parent_ancestors.iter().skip(1) {
let TypeAnnotation::CustomClass { id, .. } = anc else { unreachable!() };
anc_set.insert(id, anc.clone());
}
ancestors.extend(anc_set.into_values());
}
// push `self` as first ancestor of class
ancestors.insert(0, make_self_type_annotation(type_vars.as_slice(), *object_id));
}
}
pub fn parse_parameter_default_value(
@ -1118,3 +1209,23 @@ pub fn arraylike_get_ndims(unifier: &mut Unifier, ty: Type) -> u64 {
_ => 0,
}
}
/// Extract an ndarray's `ndims` [type][`Type`] in `u64`. Panic if not possible.
/// The `ndims` must only contain 1 value.
#[must_use]
pub fn extract_ndims(unifier: &Unifier, ndims_ty: Type) -> u64 {
let ndims_ty_enum = unifier.get_ty_immutable(ndims_ty);
let TypeEnum::TLiteral { values, .. } = &*ndims_ty_enum else {
panic!("ndims_ty should be a TLiteral");
};
assert_eq!(values.len(), 1, "ndims_ty TLiteral should only contain 1 value");
let ndims = values[0].clone();
u64::try_from(ndims).unwrap()
}
/// Return an ndarray's `ndims` as a typechecker [`Type`] from its `u64` value.
pub fn create_ndims(unifier: &mut Unifier, ndims: u64) -> Type {
unifier.get_fresh_literal(vec![SymbolValue::U64(ndims)], None)
}

View File

@ -6,36 +6,36 @@ use std::{
sync::Arc,
};
use super::codegen::CodeGenContext;
use super::typecheck::type_inferencer::PrimitiveStore;
use super::typecheck::typedef::{
FunSignature, FuncArg, SharedUnifier, Type, TypeEnum, Unifier, VarMap,
};
use crate::{
codegen::CodeGenerator,
symbol_resolver::{SymbolResolver, ValueEnum},
typecheck::{
type_inferencer::CodeLocation,
typedef::{CallId, TypeVarId},
},
};
use inkwell::values::BasicValueEnum;
use itertools::Itertools;
use nac3parser::ast::{self, Location, Stmt, StrRef};
use parking_lot::RwLock;
#[derive(PartialEq, Eq, PartialOrd, Ord, Clone, Copy, Hash, Debug)]
pub struct DefinitionId(pub usize);
use nac3parser::ast::{self, Expr, Location, Stmt, StrRef};
use crate::{
codegen::{CodeGenContext, CodeGenerator},
symbol_resolver::{SymbolResolver, ValueEnum},
typecheck::{
type_inferencer::{CodeLocation, PrimitiveStore},
typedef::{
CallId, FunSignature, FuncArg, SharedUnifier, Type, TypeEnum, TypeVarId, Unifier,
VarMap,
},
},
};
use composer::*;
use type_annotation::*;
pub mod builtins;
pub mod composer;
pub mod helper;
pub mod numpy;
pub mod type_annotation;
use composer::*;
use type_annotation::*;
#[cfg(test)]
mod test;
pub mod type_annotation;
#[derive(PartialEq, Eq, PartialOrd, Ord, Clone, Copy, Hash, Debug)]
pub struct DefinitionId(pub usize);
type GenCallCallback = dyn for<'ctx, 'a> Fn(
&mut CodeGenContext<'ctx, 'a>,
@ -148,6 +148,25 @@ pub enum TopLevelDef {
/// Definition location.
loc: Option<Location>,
},
Variable {
/// Qualified name of the global variable, should be unique globally.
name: String,
/// Simple name, the same as in method/function definition.
simple_name: StrRef,
/// Type of the global variable.
ty: Type,
/// The declared type of the global variable, or [`None`] if no type annotation is provided.
ty_decl: Option<Expr>,
/// Symbol resolver of the module defined the class.
resolver: Option<Arc<dyn SymbolResolver + Send + Sync>>,
/// Definition location.
loc: Option<Location>,
},
}
pub struct TopLevelContext {

View File

@ -1,11 +1,10 @@
use crate::{
toplevel::helper::PrimDef,
typecheck::{
use itertools::Itertools;
use super::helper::PrimDef;
use crate::typecheck::{
type_inferencer::PrimitiveStore,
typedef::{Type, TypeEnum, TypeVarId, Unifier, VarMap},
},
};
use itertools::Itertools;
/// Creates a `ndarray` [`Type`] with the given type arguments.
///

View File

@ -3,10 +3,10 @@ source: nac3core/src/toplevel/test.rs
expression: res_vec
---
[
"Class {\nname: \"Generic_A\",\nancestors: [\"Generic_A[V]\", \"B\"],\nfields: [\"aa\", \"a\"],\nmethods: [(\"__init__\", \"fn[[], none]\"), (\"foo\", \"fn[[b:T], none]\"), (\"fun\", \"fn[[a:int32], V]\")],\ntype_vars: [\"V\"]\n}\n",
"Function {\nname: \"Generic_A.__init__\",\nsig: \"fn[[], none]\",\nvar_id: []\n}\n",
"Function {\nname: \"Generic_A.fun\",\nsig: \"fn[[a:int32], V]\",\nvar_id: [TypeVarId(241)]\n}\n",
"Class {\nname: \"B\",\nancestors: [\"B\"],\nfields: [\"aa\"],\nmethods: [(\"__init__\", \"fn[[], none]\"), (\"foo\", \"fn[[b:T], none]\")],\ntype_vars: []\n}\n",
"Function {\nname: \"B.__init__\",\nsig: \"fn[[], none]\",\nvar_id: []\n}\n",
"Function {\nname: \"B.foo\",\nsig: \"fn[[b:T], none]\",\nvar_id: []\n}\n",
"Class {\nname: \"Generic_A\",\nancestors: [\"Generic_A[V]\", \"B\"],\nfields: [\"aa\", \"a\"],\nmethods: [(\"__init__\", \"fn[[], none]\"), (\"foo\", \"fn[[b:T], none]\"), (\"fun\", \"fn[[a:int32], V]\")],\ntype_vars: [\"V\"]\n}\n",
"Function {\nname: \"Generic_A.__init__\",\nsig: \"fn[[], none]\",\nvar_id: []\n}\n",
"Function {\nname: \"Generic_A.fun\",\nsig: \"fn[[a:int32], V]\",\nvar_id: [TypeVarId(246)]\n}\n",
]

View File

@ -7,11 +7,11 @@ expression: res_vec
"Function {\nname: \"A.__init__\",\nsig: \"fn[[], none]\",\nvar_id: []\n}\n",
"Function {\nname: \"A.fun\",\nsig: \"fn[[b:B], none]\",\nvar_id: []\n}\n",
"Function {\nname: \"A.foo\",\nsig: \"fn[[a:T, b:V], none]\",\nvar_id: [TypeVarId(249)]\n}\n",
"Class {\nname: \"B\",\nancestors: [\"B\", \"C\", \"A\"],\nfields: [\"a\"],\nmethods: [(\"__init__\", \"fn[[], none]\"), (\"fun\", \"fn[[b:B], none]\"), (\"foo\", \"fn[[a:T, b:V], none]\")],\ntype_vars: []\n}\n",
"Function {\nname: \"B.__init__\",\nsig: \"fn[[], none]\",\nvar_id: []\n}\n",
"Class {\nname: \"C\",\nancestors: [\"C\", \"A\"],\nfields: [\"a\"],\nmethods: [(\"__init__\", \"fn[[], none]\"), (\"fun\", \"fn[[b:B], none]\"), (\"foo\", \"fn[[a:T, b:V], none]\")],\ntype_vars: []\n}\n",
"Function {\nname: \"C.__init__\",\nsig: \"fn[[], none]\",\nvar_id: []\n}\n",
"Function {\nname: \"C.fun\",\nsig: \"fn[[b:B], none]\",\nvar_id: []\n}\n",
"Class {\nname: \"B\",\nancestors: [\"B\", \"C\", \"A\"],\nfields: [\"a\"],\nmethods: [(\"__init__\", \"fn[[], none]\"), (\"fun\", \"fn[[b:B], none]\"), (\"foo\", \"fn[[a:T, b:V], none]\")],\ntype_vars: []\n}\n",
"Function {\nname: \"B.__init__\",\nsig: \"fn[[], none]\",\nvar_id: []\n}\n",
"Function {\nname: \"foo\",\nsig: \"fn[[a:A], none]\",\nvar_id: []\n}\n",
"Function {\nname: \"ff\",\nsig: \"fn[[a:T], V]\",\nvar_id: [TypeVarId(257)]\n}\n",
]

View File

@ -1,21 +1,23 @@
use super::*;
use crate::toplevel::helper::PrimDef;
use crate::typecheck::typedef::into_var_map;
use std::{collections::HashMap, sync::Arc};
use indoc::indoc;
use parking_lot::Mutex;
use test_case::test_case;
use nac3parser::{
ast::{fold::Fold, FileName},
parser::parse_program,
};
use super::{helper::PrimDef, DefinitionId, *};
use crate::{
codegen::CodeGenContext,
symbol_resolver::{SymbolResolver, ValueEnum},
toplevel::DefinitionId,
typecheck::{
type_inferencer::PrimitiveStore,
typedef::{Type, Unifier},
typedef::{into_var_map, Type, Unifier},
},
};
use indoc::indoc;
use nac3parser::ast::FileName;
use nac3parser::{ast::fold::Fold, parser::parse_program};
use parking_lot::Mutex;
use std::{collections::HashMap, sync::Arc};
use test_case::test_case;
struct ResolverInternal {
id_to_type: Mutex<HashMap<StrRef, Type>>,
@ -62,6 +64,7 @@ impl SymbolResolver for Resolver {
&self,
_: StrRef,
_: &mut CodeGenContext<'ctx, '_>,
_: &mut dyn CodeGenerator,
) -> Option<ValueEnum<'ctx>> {
unimplemented!()
}
@ -226,11 +229,6 @@ fn test_simple_function_analyze(source: &[&str], tys: &[&str], names: &[&str]) {
def foo(self, a: T, b: V):
pass
"},
indoc! {"
class B(C):
def __init__(self):
pass
"},
indoc! {"
class C(A):
def __init__(self):
@ -239,6 +237,11 @@ fn test_simple_function_analyze(source: &[&str], tys: &[&str], names: &[&str]) {
a = 1
pass
"},
indoc! {"
class B(C):
def __init__(self):
pass
"},
indoc! {"
def foo(a: A):
pass
@ -253,14 +256,6 @@ fn test_simple_function_analyze(source: &[&str], tys: &[&str], names: &[&str]) {
)]
#[test_case(
&[
indoc! {"
class Generic_A(Generic[V], B):
a: int64
def __init__(self):
self.a = 123123123123
def fun(self, a: int32) -> V:
pass
"},
indoc! {"
class B:
aa: bool
@ -268,6 +263,14 @@ fn test_simple_function_analyze(source: &[&str], tys: &[&str], names: &[&str]) {
self.aa = False
def foo(self, b: T):
pass
"},
indoc! {"
class Generic_A(Generic[V], B):
a: int64
def __init__(self):
self.a = 123123123123
def fun(self, a: int32) -> V:
pass
"}
],
&[];
@ -387,18 +390,18 @@ fn test_simple_function_analyze(source: &[&str], tys: &[&str], names: &[&str]) {
pass
"}
],
&["cyclic inheritance detected"];
&["NameError: name 'B' is not defined (at unknown:1:9)"];
"cyclic1"
)]
#[test_case(
&[
indoc! {"
class A(B[bool, int64]):
class B(Generic[V, T], C[int32]):
def __init__(self):
pass
"},
indoc! {"
class B(Generic[V, T], C[int32]):
class A(B[bool, int64]):
def __init__(self):
pass
"},
@ -408,7 +411,7 @@ fn test_simple_function_analyze(source: &[&str], tys: &[&str], names: &[&str]) {
pass
"},
],
&["cyclic inheritance detected"];
&["NameError: name 'C' is not defined (at unknown:1:25)"];
"cyclic2"
)]
#[test_case(
@ -432,11 +435,6 @@ fn test_simple_function_analyze(source: &[&str], tys: &[&str], names: &[&str]) {
)]
#[test_case(
&[
indoc! {"
class A(B, Generic[T], C):
def __init__(self):
pass
"},
indoc! {"
class B:
def __init__(self):
@ -446,6 +444,11 @@ fn test_simple_function_analyze(source: &[&str], tys: &[&str], names: &[&str]) {
class C:
def __init__(self):
pass
"},
indoc! {"
class A(B, Generic[T], C):
def __init__(self):
pass
"}
],

View File

@ -1,10 +1,13 @@
use super::*;
use crate::symbol_resolver::SymbolValue;
use crate::toplevel::helper::{PrimDef, PrimDefDetails};
use crate::typecheck::typedef::VarMap;
use nac3parser::ast::Constant;
use strum::IntoEnumIterator;
use nac3parser::ast::Constant;
use super::{
helper::{PrimDef, PrimDefDetails},
*,
};
use crate::{symbol_resolver::SymbolValue, typecheck::typedef::VarMap};
#[derive(Clone, Debug)]
pub enum TypeAnnotation {
Primitive(Type),
@ -97,7 +100,13 @@ pub fn parse_ast_to_type_annotation_kinds<T, S: std::hash::BuildHasher + Clone>(
Ok(TypeAnnotation::CustomClass { id: PrimDef::Exception.id(), params: Vec::default() })
} else if let Ok(obj_id) = resolver.get_identifier_def(*id) {
let type_vars = {
let def_read = top_level_defs[obj_id.0].try_read();
let Some(top_level_def) = top_level_defs.get(obj_id.0) else {
return Err(HashSet::from([format!(
"NameError: name '{id}' is not defined (at {})",
expr.location
)]));
};
let def_read = top_level_def.try_read();
if let Some(def_read) = def_read {
if let TopLevelDef::Class { type_vars, .. } = &*def_read {
type_vars.clone()
@ -152,12 +161,17 @@ pub fn parse_ast_to_type_annotation_kinds<T, S: std::hash::BuildHasher + Clone>(
}
let obj_id = resolver.get_identifier_def(*id)?;
let type_vars = {
let def_read = top_level_defs[obj_id.0].try_read();
let Some(top_level_def) = top_level_defs.get(obj_id.0) else {
return Err(HashSet::from([format!(
"NameError: name '{id}' is not defined (at {})",
expr.location
)]));
};
let def_read = top_level_def.try_read();
if let Some(def_read) = def_read {
let TopLevelDef::Class { type_vars, .. } = &*def_read else {
unreachable!("must be class here")
};
type_vars.clone()
} else {
locked.get(&obj_id).unwrap().clone()

View File

@ -1,13 +1,19 @@
use crate::toplevel::helper::PrimDef;
use std::{
collections::{HashMap, HashSet},
iter::once,
};
use super::type_inferencer::Inferencer;
use super::typedef::{Type, TypeEnum};
use nac3parser::ast::{
self, Constant, Expr, ExprKind,
Operator::{LShift, RShift},
Stmt, StmtKind, StrRef,
};
use std::{collections::HashSet, iter::once};
use super::{
type_inferencer::{DeclarationSource, IdentifierInfo, Inferencer},
typedef::{Type, TypeEnum},
};
use crate::toplevel::helper::PrimDef;
impl<'a> Inferencer<'a> {
fn should_have_value(&mut self, expr: &Expr<Option<Type>>) -> Result<(), HashSet<String>> {
@ -21,15 +27,29 @@ impl<'a> Inferencer<'a> {
fn check_pattern(
&mut self,
pattern: &Expr<Option<Type>>,
defined_identifiers: &mut HashSet<StrRef>,
defined_identifiers: &mut HashMap<StrRef, IdentifierInfo>,
) -> Result<(), HashSet<String>> {
match &pattern.node {
ExprKind::Name { id, .. } if id == &"none".into() => {
Err(HashSet::from([format!("cannot assign to a `none` (at {})", pattern.location)]))
}
ExprKind::Name { id, .. } => {
if !defined_identifiers.contains(id) {
defined_identifiers.insert(*id);
// If `id` refers to a declared symbol, reject this assignment if it is used in the
// context of an (implicit) global variable
if let Some(id_info) = defined_identifiers.get(id) {
if matches!(
id_info.source,
DeclarationSource::Global { is_explicit: Some(false) }
) {
return Err(HashSet::from([format!(
"cannot access local variable '{id}' before it is declared (at {})",
pattern.location
)]));
}
}
if !defined_identifiers.contains_key(id) {
defined_identifiers.insert(*id, IdentifierInfo::default());
}
self.should_have_value(pattern)?;
Ok(())
@ -69,7 +89,7 @@ impl<'a> Inferencer<'a> {
fn check_expr(
&mut self,
expr: &Expr<Option<Type>>,
defined_identifiers: &mut HashSet<StrRef>,
defined_identifiers: &mut HashMap<StrRef, IdentifierInfo>,
) -> Result<(), HashSet<String>> {
// there are some cases where the custom field is None
if let Some(ty) = &expr.custom {
@ -90,7 +110,7 @@ impl<'a> Inferencer<'a> {
return Ok(());
}
self.should_have_value(expr)?;
if !defined_identifiers.contains(id) {
if !defined_identifiers.contains_key(id) {
match self.function_data.resolver.get_symbol_type(
self.unifier,
&self.top_level.definitions.read(),
@ -98,7 +118,22 @@ impl<'a> Inferencer<'a> {
*id,
) {
Ok(_) => {
self.defined_identifiers.insert(*id);
let is_global = self.is_id_global(*id);
defined_identifiers.insert(
*id,
IdentifierInfo {
source: match is_global {
Some(true) => {
DeclarationSource::Global { is_explicit: Some(false) }
}
Some(false) => {
DeclarationSource::Global { is_explicit: None }
}
None => DeclarationSource::Local,
},
},
);
}
Err(e) => {
return Err(HashSet::from([format!(
@ -171,9 +206,7 @@ impl<'a> Inferencer<'a> {
let mut defined_identifiers = defined_identifiers.clone();
for arg in &args.args {
// TODO: should we check the types here?
if !defined_identifiers.contains(&arg.node.arg) {
defined_identifiers.insert(arg.node.arg);
}
defined_identifiers.entry(arg.node.arg).or_default();
}
self.check_expr(body, &mut defined_identifiers)?;
}
@ -236,7 +269,7 @@ impl<'a> Inferencer<'a> {
fn check_stmt(
&mut self,
stmt: &Stmt<Option<Type>>,
defined_identifiers: &mut HashSet<StrRef>,
defined_identifiers: &mut HashMap<StrRef, IdentifierInfo>,
) -> Result<bool, HashSet<String>> {
match &stmt.node {
StmtKind::For { target, iter, body, orelse, .. } => {
@ -262,9 +295,11 @@ impl<'a> Inferencer<'a> {
let body_returned = self.check_block(body, &mut body_identifiers)?;
let orelse_returned = self.check_block(orelse, &mut orelse_identifiers)?;
for ident in &body_identifiers {
if !defined_identifiers.contains(ident) && orelse_identifiers.contains(ident) {
defined_identifiers.insert(*ident);
for ident in body_identifiers.keys() {
if !defined_identifiers.contains_key(ident)
&& orelse_identifiers.contains_key(ident)
{
defined_identifiers.insert(*ident, IdentifierInfo::default());
}
}
Ok(body_returned && orelse_returned)
@ -295,7 +330,7 @@ impl<'a> Inferencer<'a> {
let mut defined_identifiers = defined_identifiers.clone();
let ast::ExcepthandlerKind::ExceptHandler { name, body, .. } = &handler.node;
if let Some(name) = name {
defined_identifiers.insert(*name);
defined_identifiers.insert(*name, IdentifierInfo::default());
}
self.check_block(body, &mut defined_identifiers)?;
}
@ -359,6 +394,44 @@ impl<'a> Inferencer<'a> {
}
Ok(true)
}
StmtKind::Global { names, .. } => {
for id in names {
if let Some(id_info) = defined_identifiers.get(id) {
if id_info.source == DeclarationSource::Local {
return Err(HashSet::from([format!(
"name '{id}' is referenced prior to global declaration at {}",
stmt.location,
)]));
}
continue;
}
match self.function_data.resolver.get_symbol_type(
self.unifier,
&self.top_level.definitions.read(),
self.primitives,
*id,
) {
Ok(_) => {
defined_identifiers.insert(
*id,
IdentifierInfo {
source: DeclarationSource::Global { is_explicit: Some(true) },
},
);
}
Err(e) => {
return Err(HashSet::from([format!(
"type error at identifier `{}` ({}) at {}",
id, e, stmt.location
)]))
}
}
}
Ok(false)
}
// break, raise, etc.
_ => Ok(false),
}
@ -367,7 +440,7 @@ impl<'a> Inferencer<'a> {
pub fn check_block(
&mut self,
block: &[Stmt<Option<Type>>],
defined_identifiers: &mut HashSet<StrRef>,
defined_identifiers: &mut HashMap<StrRef, IdentifierInfo>,
) -> Result<bool, HashSet<String>> {
let mut ret = false;
for stmt in block {

View File

@ -1,17 +1,21 @@
use crate::symbol_resolver::SymbolValue;
use crate::toplevel::helper::PrimDef;
use crate::toplevel::numpy::{make_ndarray_ty, unpack_ndarray_var_tys};
use crate::typecheck::{
use std::{cmp::max, collections::HashMap, rc::Rc};
use itertools::{iproduct, Itertools};
use strum::IntoEnumIterator;
use nac3parser::ast::{Cmpop, Operator, StrRef, Unaryop};
use super::{
type_inferencer::*,
typedef::{FunSignature, FuncArg, Type, TypeEnum, Unifier, VarMap},
};
use itertools::{iproduct, Itertools};
use nac3parser::ast::StrRef;
use nac3parser::ast::{Cmpop, Operator, Unaryop};
use std::cmp::max;
use std::collections::HashMap;
use std::rc::Rc;
use strum::IntoEnumIterator;
use crate::{
symbol_resolver::SymbolValue,
toplevel::{
helper::PrimDef,
numpy::{make_ndarray_ty, unpack_ndarray_var_tys},
},
};
/// The variant of a binary operator.
#[derive(Debug, Clone, Copy, PartialEq, Eq)]

View File

@ -1,14 +1,13 @@
use std::collections::HashMap;
use std::fmt::Display;
use std::{collections::HashMap, fmt::Display};
use crate::typecheck::{magic_methods::HasOpInfo, typedef::TypeEnum};
use itertools::Itertools;
use nac3parser::ast::{Cmpop, Location, StrRef};
use super::{
magic_methods::Binop,
typedef::{RecordKey, Type, Unifier},
magic_methods::{Binop, HasOpInfo},
typedef::{RecordKey, Type, TypeEnum, Unifier},
};
use itertools::Itertools;
use nac3parser::ast::{Cmpop, Location, StrRef};
#[derive(Debug, Clone)]
pub enum TypeErrorKind {

View File

@ -1,32 +1,36 @@
use std::cmp::max;
use std::collections::{HashMap, HashSet};
use std::convert::{From, TryInto};
use std::iter::once;
use std::{cell::RefCell, sync::Arc};
use std::{
cell::RefCell,
cmp::max,
collections::{HashMap, HashSet},
convert::{From, TryInto},
iter::once,
sync::Arc,
};
use itertools::{izip, Itertools};
use nac3parser::ast::{
self,
fold::{self, Fold},
Arguments, Comprehension, ExprContext, ExprKind, Ident, Located, Location, StrRef,
};
use super::{
magic_methods::*,
type_error::{TypeError, TypeErrorKind},
typedef::{
into_var_map, iter_type_vars, Call, CallId, FunSignature, FuncArg, OperatorInfo,
into_var_map, iter_type_vars, Call, CallId, FunSignature, FuncArg, Mapping, OperatorInfo,
RecordField, RecordKey, Type, TypeEnum, TypeVar, Unifier, VarMap,
},
};
use crate::toplevel::type_annotation::TypeAnnotation;
use crate::{
symbol_resolver::{SymbolResolver, SymbolValue},
toplevel::{
helper::{arraylike_flatten_element_type, arraylike_get_ndims, PrimDef},
numpy::{make_ndarray_ty, unpack_ndarray_var_tys},
type_annotation::TypeAnnotation,
TopLevelContext, TopLevelDef,
},
typecheck::typedef::Mapping,
};
use itertools::{izip, Itertools};
use nac3parser::ast::{
self,
fold::{self, Fold},
Arguments, Comprehension, ExprContext, ExprKind, Located, Location, StrRef,
};
#[cfg(test)]
@ -84,6 +88,40 @@ impl PrimitiveStore {
}
}
/// The location where an identifier declaration refers to.
#[derive(Clone, Copy, Debug, Eq, PartialEq)]
pub enum DeclarationSource {
/// Local scope.
Local,
/// Global scope.
Global {
/// Whether the identifier is declared by the use of `global` statement. This field is
/// [`None`] if the identifier does not refer to a variable.
is_explicit: Option<bool>,
},
}
/// Information regarding a defined identifier.
#[derive(Clone, Copy, Debug)]
pub struct IdentifierInfo {
/// Whether this identifier refers to a global variable.
pub source: DeclarationSource,
}
impl Default for IdentifierInfo {
fn default() -> Self {
IdentifierInfo { source: DeclarationSource::Local }
}
}
impl IdentifierInfo {
#[must_use]
pub fn new() -> IdentifierInfo {
IdentifierInfo::default()
}
}
pub struct FunctionData {
pub resolver: Arc<dyn SymbolResolver + Send + Sync>,
pub return_type: Option<Type>,
@ -92,7 +130,7 @@ pub struct FunctionData {
pub struct Inferencer<'a> {
pub top_level: &'a TopLevelContext,
pub defined_identifiers: HashSet<StrRef>,
pub defined_identifiers: HashMap<StrRef, IdentifierInfo>,
pub function_data: &'a mut FunctionData,
pub unifier: &'a mut Unifier,
pub primitives: &'a PrimitiveStore,
@ -224,9 +262,7 @@ impl<'a> Fold<()> for Inferencer<'a> {
handler.location,
));
if let Some(name) = name {
if !self.defined_identifiers.contains(&name) {
self.defined_identifiers.insert(name);
}
self.defined_identifiers.entry(name).or_default();
if let Some(old_typ) = self.variable_mapping.insert(name, typ) {
let loc = handler.location;
self.unifier.unify(old_typ, typ).map_err(|e| {
@ -378,6 +414,7 @@ impl<'a> Fold<()> for Inferencer<'a> {
| ast::StmtKind::Continue { .. }
| ast::StmtKind::Expr { .. }
| ast::StmtKind::For { .. }
| ast::StmtKind::Global { .. }
| ast::StmtKind::Pass { .. }
| ast::StmtKind::Try { .. } => {}
ast::StmtKind::If { test, .. } | ast::StmtKind::While { test, .. } => {
@ -499,9 +536,8 @@ impl<'a> Fold<()> for Inferencer<'a> {
}
ast::StmtKind::Assert { test, msg, .. } => {
self.unify(test.custom.unwrap(), self.primitives.bool, &test.location)?;
match msg {
Some(m) => self.unify(m.custom.unwrap(), self.primitives.str, &m.location)?,
None => (),
if let Some(m) = msg {
self.unify(m.custom.unwrap(), self.primitives.str, &m.location)?;
}
}
_ => return report_error("Unsupported statement type", stmt.location),
@ -549,7 +585,7 @@ impl<'a> Fold<()> for Inferencer<'a> {
unreachable!("must be tobj")
}
} else {
if !self.defined_identifiers.contains(id) {
if !self.defined_identifiers.contains_key(id) {
match self.function_data.resolver.get_symbol_type(
self.unifier,
&self.top_level.definitions.read(),
@ -557,7 +593,22 @@ impl<'a> Fold<()> for Inferencer<'a> {
*id,
) {
Ok(_) => {
self.defined_identifiers.insert(*id);
let is_global = self.is_id_global(*id);
self.defined_identifiers.insert(
*id,
IdentifierInfo {
source: match is_global {
Some(true) => DeclarationSource::Global {
is_explicit: Some(false),
},
Some(false) => {
DeclarationSource::Global { is_explicit: None }
}
None => DeclarationSource::Local,
},
},
);
}
Err(e) => {
return report_error(
@ -622,8 +673,8 @@ impl<'a> Inferencer<'a> {
fn infer_pattern<T>(&mut self, pattern: &ast::Expr<T>) -> Result<(), InferenceError> {
match &pattern.node {
ExprKind::Name { id, .. } => {
if !self.defined_identifiers.contains(id) {
self.defined_identifiers.insert(*id);
if !self.defined_identifiers.contains_key(id) {
self.defined_identifiers.insert(*id, IdentifierInfo::default());
}
Ok(())
}
@ -732,8 +783,8 @@ impl<'a> Inferencer<'a> {
let mut defined_identifiers = self.defined_identifiers.clone();
for arg in &args.args {
let name = &arg.node.arg;
if !defined_identifiers.contains(name) {
defined_identifiers.insert(*name);
if !defined_identifiers.contains_key(name) {
defined_identifiers.insert(*name, IdentifierInfo::default());
}
}
let fn_args: Vec<_> = args
@ -1550,36 +1601,29 @@ impl<'a> Inferencer<'a> {
}
// 2-argument ndarray n-dimensional creation functions
if id == &"np_full".into() && args.len() == 2 {
let ExprKind::List { elts, .. } = &args[0].node else {
return report_error(
format!(
"Expected List literal for first argument of {id}, got {}",
args[0].node.name()
)
.as_str(),
args[0].location,
);
};
// Parse arguments
let shape_expr = args.remove(0);
let (ndims, shape) =
self.fold_numpy_function_call_shape_argument(*id, 0, shape_expr)?; // Special handling for `shape`
let ndims = elts.len() as u64;
let fill_value = self.fold_expr(args.remove(0))?;
let arg0 = self.fold_expr(args.remove(0))?;
let arg1 = self.fold_expr(args.remove(0))?;
let ty = arg1.custom.unwrap();
// Build the return type
let dtype = fill_value.custom.unwrap();
let ndims = self.unifier.get_fresh_literal(vec![SymbolValue::U64(ndims)], None);
let ret = make_ndarray_ty(self.unifier, self.primitives, Some(ty), Some(ndims));
let ret = make_ndarray_ty(self.unifier, self.primitives, Some(dtype), Some(ndims));
let custom = self.unifier.add_ty(TypeEnum::TFunc(FunSignature {
args: vec![
FuncArg {
name: "shape".into(),
ty: arg0.custom.unwrap(),
ty: shape.custom.unwrap(),
default_value: None,
is_vararg: false,
},
FuncArg {
name: "fill_value".into(),
ty: arg1.custom.unwrap(),
ty: fill_value.custom.unwrap(),
default_value: None,
is_vararg: false,
},
@ -1597,7 +1641,7 @@ impl<'a> Inferencer<'a> {
location: func.location,
node: ExprKind::Name { id: *id, ctx: *ctx },
}),
args: vec![arg0, arg1],
args: vec![shape, fill_value],
keywords: vec![],
},
}));
@ -2640,4 +2684,22 @@ impl<'a> Inferencer<'a> {
self.constrain(body.custom.unwrap(), orelse.custom.unwrap(), &body.location)?;
Ok(body.custom.unwrap())
}
/// Determines whether the given `id` refers to a global symbol.
///
/// Returns `Some(true)` if `id` refers to a global variable, `Some(false)` if `id` refers to a
/// class/function, and `None` if `id` refers to a local symbol.
pub(super) fn is_id_global(&self, id: Ident) -> Option<bool> {
self.top_level
.definitions
.read()
.iter()
.map(|def| match *def.read() {
TopLevelDef::Class { name, .. } => (name, false),
TopLevelDef::Function { simple_name, .. } => (simple_name, false),
TopLevelDef::Variable { simple_name, .. } => (simple_name, true),
})
.find(|(global, _)| global == &id)
.map(|(_, has_explicit_prop)| has_explicit_prop)
}
}

View File

@ -1,18 +1,20 @@
use super::super::{magic_methods::with_fields, typedef::*};
use super::*;
use crate::{
codegen::CodeGenContext,
symbol_resolver::ValueEnum,
toplevel::{helper::PrimDef, DefinitionId, TopLevelDef},
};
use std::iter::zip;
use indexmap::IndexMap;
use indoc::indoc;
use nac3parser::ast::FileName;
use nac3parser::parser::parse_program;
use parking_lot::RwLock;
use std::iter::zip;
use test_case::test_case;
use nac3parser::{ast::FileName, parser::parse_program};
use super::*;
use crate::{
codegen::{CodeGenContext, CodeGenerator},
symbol_resolver::ValueEnum,
toplevel::{helper::PrimDef, DefinitionId, TopLevelDef},
typecheck::{magic_methods::with_fields, typedef::*},
};
struct Resolver {
id_to_type: HashMap<StrRef, Type>,
id_to_def: HashMap<StrRef, DefinitionId>,
@ -41,6 +43,7 @@ impl SymbolResolver for Resolver {
&self,
_: StrRef,
_: &mut CodeGenContext<'ctx, '_>,
_: &mut dyn CodeGenerator,
) -> Option<ValueEnum<'ctx>> {
unimplemented!()
}
@ -517,7 +520,7 @@ impl TestEnvironment {
primitives: &mut self.primitives,
virtual_checks: &mut self.virtual_checks,
calls: &mut self.calls,
defined_identifiers: HashSet::default(),
defined_identifiers: HashMap::default(),
in_handler: false,
}
}
@ -593,8 +596,9 @@ fn test_basic(source: &str, mapping: &HashMap<&str, &str>, virtuals: &[(&str, &s
println!("source:\n{source}");
let mut env = TestEnvironment::new();
let id_to_name = std::mem::take(&mut env.id_to_name);
let mut defined_identifiers: HashSet<_> = env.identifier_mapping.keys().copied().collect();
defined_identifiers.insert("virtual".into());
let mut defined_identifiers: HashMap<_, _> =
env.identifier_mapping.keys().copied().map(|id| (id, IdentifierInfo::default())).collect();
defined_identifiers.insert("virtual".into(), IdentifierInfo::default());
let mut inferencer = env.get_inferencer();
inferencer.defined_identifiers.clone_from(&defined_identifiers);
let statements = parse_program(source, FileName::default()).unwrap();
@ -739,8 +743,9 @@ fn test_primitive_magic_methods(source: &str, mapping: &HashMap<&str, &str>) {
println!("source:\n{source}");
let mut env = TestEnvironment::basic_test_env();
let id_to_name = std::mem::take(&mut env.id_to_name);
let mut defined_identifiers: HashSet<_> = env.identifier_mapping.keys().copied().collect();
defined_identifiers.insert("virtual".into());
let mut defined_identifiers: HashMap<_, _> =
env.identifier_mapping.keys().copied().map(|id| (id, IdentifierInfo::default())).collect();
defined_identifiers.insert("virtual".into(), IdentifierInfo::default());
let mut inferencer = env.get_inferencer();
inferencer.defined_identifiers.clone_from(&defined_identifiers);
let statements = parse_program(source, FileName::default()).unwrap();

View File

@ -1,21 +1,28 @@
use super::magic_methods::{Binop, HasOpInfo};
use super::type_error::{TypeError, TypeErrorKind};
use super::unification_table::{UnificationKey, UnificationTable};
use crate::symbol_resolver::SymbolValue;
use crate::toplevel::helper::PrimDef;
use crate::toplevel::{DefinitionId, TopLevelContext, TopLevelDef};
use crate::typecheck::magic_methods::OpInfo;
use crate::typecheck::type_inferencer::PrimitiveStore;
use std::{
borrow::Cow,
cell::RefCell,
collections::{HashMap, HashSet},
fmt::{self, Display},
iter::{repeat, zip},
rc::Rc,
sync::{Arc, Mutex},
};
use indexmap::IndexMap;
use itertools::{repeat_n, Itertools};
use nac3parser::ast::{Cmpop, Location, StrRef, Unaryop};
use std::cell::RefCell;
use std::collections::HashMap;
use std::fmt::{self, Display};
use std::iter::{repeat, zip};
use std::rc::Rc;
use std::sync::{Arc, Mutex};
use std::{borrow::Cow, collections::HashSet};
use super::{
magic_methods::{Binop, HasOpInfo, OpInfo},
type_error::{TypeError, TypeErrorKind},
type_inferencer::PrimitiveStore,
unification_table::{UnificationKey, UnificationTable},
};
use crate::{
symbol_resolver::SymbolValue,
toplevel::{helper::PrimDef, DefinitionId, TopLevelContext, TopLevelDef},
};
#[cfg(test)]
mod test;
@ -670,8 +677,8 @@ impl Unifier {
let num_args = posargs.len() + kwargs.len();
// Now we check the arguments against the parameters,
// and depending on what `call_info` is, we might change how the behavior `unify_call()`
// in hopes to improve user error messages when type checking fails.
// and depending on what `call_info` is, we might change how `unify_call()` behaves
// to improve user error messages when type checking fails.
match operator_info {
Some(OperatorInfo::IsBinaryOp { self_type, operator }) => {
// The call is written in the form of (say) `a + b`.

View File

@ -1,10 +1,12 @@
use super::super::magic_methods::with_fields;
use super::*;
use std::collections::HashMap;
use indoc::indoc;
use itertools::Itertools;
use std::collections::HashMap;
use test_case::test_case;
use super::*;
use crate::typecheck::magic_methods::with_fields;
impl Unifier {
/// Check whether two types are equal.
fn eq(&mut self, a: Type, b: Type) -> bool {

View File

@ -24,26 +24,15 @@ pub const DW_EH_PE_aligned: u8 = 0x50;
pub const DW_EH_PE_indirect: u8 = 0x80;
#[derive(Clone)]
pub struct DwarfReader<'a> {
pub slice: &'a [u8],
pub virt_addr: u32,
base_slice: &'a [u8],
base_virt_addr: u32,
}
impl<'a> DwarfReader<'a> {
pub fn new(slice: &[u8], virt_addr: u32) -> DwarfReader {
DwarfReader { slice, virt_addr, base_slice: slice, base_virt_addr: virt_addr }
}
/// Creates a new instance from another instance of [DwarfReader], optionally removing any
/// offsets previously applied to the other instance.
pub fn from_reader(other: &DwarfReader<'a>, reset_offset: bool) -> DwarfReader<'a> {
if reset_offset {
DwarfReader::new(other.base_slice, other.base_virt_addr)
} else {
DwarfReader::new(other.slice, other.virt_addr)
}
DwarfReader { slice, virt_addr }
}
pub fn offset(&mut self, offset: u32) {
@ -90,6 +79,7 @@ impl<'a> DwarfReader<'a> {
pub fn read_u8(&mut self) -> u8 {
let val = self.slice[0];
self.slice = &self.slice[1..];
self.virt_addr += 1;
val
}
}
@ -101,6 +91,7 @@ macro_rules! impl_read_fn {
pub fn $byteorder_fn(&mut self) -> $type {
let val = LittleEndian::$byteorder_fn(self.slice);
self.slice = &self.slice[mem::size_of::<$type>()..];
self.virt_addr += mem::size_of::<$type>() as u32;
val
}
)*
@ -226,7 +217,7 @@ impl<'a> EH_Frame<'a> {
/// Returns an [Iterator] over all Call Frame Information (CFI) records.
pub fn cfi_records(&self) -> CFI_Records<'a> {
let reader = DwarfReader::from_reader(&self.reader, true);
let reader = self.reader.clone();
let len = reader.slice.len();
CFI_Records { reader, available: len }
@ -257,7 +248,7 @@ impl<'a> CFI_Record<'a> {
0xFFFF_FFFF => unimplemented!(),
_ => {
let mut fde_reader = DwarfReader::from_reader(cie_reader, false);
let mut fde_reader = cie_reader.clone();
fde_reader.offset(length);
fde_reader
}
@ -276,7 +267,7 @@ impl<'a> CFI_Record<'a> {
// Skip code/data alignment factors & return address register along the way as well
// We only tackle the case where 'z' and 'R' are part of the augmentation string, otherwise
// we cannot get the addresses to make .eh_frame_hdr
let mut aug_data_reader = DwarfReader::from_reader(cie_reader, false);
let mut aug_data_reader = cie_reader.clone();
let mut aug_str_len = 0;
loop {
if aug_data_reader.read_u8() == b'\0' {
@ -319,7 +310,7 @@ impl<'a> CFI_Record<'a> {
/// Returns a [DwarfReader] initialized to the first Frame Description Entry (FDE) of this CFI
/// record.
pub fn get_fde_reader(&self) -> DwarfReader<'a> {
DwarfReader::from_reader(&self.fde_reader, true)
self.fde_reader.clone()
}
/// Returns an [Iterator] over all Frame Description Entries (FDEs).
@ -347,7 +338,7 @@ impl<'a> Iterator for CFI_Records<'a> {
return None;
}
let mut this_reader = DwarfReader::from_reader(&self.reader, false);
let mut this_reader = self.reader.clone();
// Remove the length of the header and the content from the counter
let length = self.reader.read_u32();
@ -360,7 +351,7 @@ impl<'a> Iterator for CFI_Records<'a> {
// Remove the length of the header and the content from the counter
self.available -= length + mem::size_of::<u32>();
let mut next_reader = DwarfReader::from_reader(&self.reader, false);
let mut next_reader = self.reader.clone();
next_reader.offset(length as u32);
let cie_ptr = self.reader.read_u32();
@ -396,6 +387,8 @@ impl<'a> Iterator for FDE_Records<'a> {
return None;
}
let fde_addr = self.reader.virt_addr;
// Remove the length of the header and the content from the counter
let length = match self.reader.read_u32() {
// eh_frame with 0-length means the CIE is terminated
@ -406,14 +399,14 @@ impl<'a> Iterator for FDE_Records<'a> {
// Remove the length of the header and the content from the counter
self.available -= length + mem::size_of::<u32>();
let mut next_fde_reader = DwarfReader::from_reader(&self.reader, false);
let mut next_fde_reader = self.reader.clone();
next_fde_reader.offset(length as u32);
let cie_ptr = self.reader.read_u32();
let next_val = if cie_ptr != 0 {
let pc_begin = read_encoded_pointer_with_pc(&mut self.reader, self.pointer_encoding)
.expect("Failed to read PC Begin");
Some((pc_begin as u32, self.reader.virt_addr))
Some((pc_begin as u32, fde_addr))
} else {
None
};
@ -446,9 +439,7 @@ impl<'a> EH_Frame_Hdr<'a> {
writer.write_u8(0x03); // fde_count_enc - 4-byte unsigned value
writer.write_u8(0x3B); // table_enc - .eh_frame_hdr section-relative 4-byte signed value
let eh_frame_offset = eh_frame_addr.wrapping_sub(
eh_frame_hdr_addr + writer.offset as u32 + ((mem::size_of::<u8>() as u32) * 4),
);
let eh_frame_offset = eh_frame_addr.wrapping_sub(eh_frame_hdr_addr + writer.offset as u32);
writer.write_u32(eh_frame_offset); // eh_frame_ptr
writer.write_u32(0); // `fde_count`, will be written in finalize_fde

View File

@ -1,10 +1,4 @@
#![deny(
future_incompatible,
let_underscore,
nonstandard_style,
rust_2024_compatibility,
clippy::all
)]
#![deny(future_incompatible, let_underscore, nonstandard_style, clippy::all)]
#![warn(clippy::pedantic)]
#![allow(
clippy::cast_possible_truncation,
@ -21,13 +15,12 @@
clippy::wildcard_imports
)]
use std::{collections::HashMap, mem, ptr, slice, str};
use byteorder::{ByteOrder, LittleEndian};
use dwarf::*;
use elf::*;
use std::collections::HashMap;
use std::{mem, ptr, slice, str};
extern crate byteorder;
use byteorder::{ByteOrder, LittleEndian};
mod dwarf;
mod elf;

Some files were not shown because too many files have changed in this diff Show More