Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Improves the error message of invalid index expression use #696

Merged
merged 3 commits into from
Aug 28, 2024
Merged

Improves the error message of invalid index expression use #696

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
6b17376
Improves the error message of invalid index expression use
rtzam Aug 23, 2024
c748b38
🤖 apply linter changes (will not trigger CI)
rtzam Aug 23, 2024
2dc8105
Merge branch 'main' into improves-error-messaging-for-index-exprs
yamaguchi1024 Aug 28, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
71 changes: 36 additions & 35 deletions src/exo/pyparser.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -128,6 +128,24 @@ def getsrcinfo(node):
# Parser Pass object


# detect which sort of type we have here
_is_size = lambda x: isinstance(x, pyast.Name) and x.id == "size"
_is_index = lambda x: isinstance(x, pyast.Name) and x.id == "index"
_is_bool = lambda x: isinstance(x, pyast.Name) and x.id == "bool"
_is_stride = lambda x: isinstance(x, pyast.Name) and x.id == "stride"

_prim_types = {
"R": UAST.Num(),
"f16": UAST.F16(),
"f32": UAST.F32(),
"f64": UAST.F64(),
"i8": UAST.INT8(),
"ui8": UAST.UINT8(),
"ui16": UAST.UINT16(),
"i32": UAST.INT32(),
}


class Parser:
def __init__(
self,
Expand Down Expand Up @@ -312,8 +330,8 @@ def parse_config_field(self, stmt):
"index": UAST.Index(),
"stride": UAST.Stride(),
}
for k in Parser._prim_types:
typ_list[k] = Parser._prim_types[k]
for k in _prim_types:
typ_list[k] = _prim_types[k]
del typ_list["R"]

if (
Expand Down Expand Up @@ -348,35 +366,29 @@ def is_at(x):
typ_node = node
mem_node = None

# detect which sort of type we have here
is_size = lambda x: isinstance(x, pyast.Name) and x.id == "size"
is_index = lambda x: isinstance(x, pyast.Name) and x.id == "index"
is_bool = lambda x: isinstance(x, pyast.Name) and x.id == "bool"
is_stride = lambda x: isinstance(x, pyast.Name) and x.id == "stride"

# parse each kind of type here
if is_size(typ_node):
if _is_size(typ_node):
if mem_node is not None:
self.err(
node, "size types should not be annotated with " "memory locations"
)
return UAST.Size(), None

elif is_index(typ_node):
elif _is_index(typ_node):
if mem_node is not None:
self.err(
node, "size types should not be annotated with " "memory locations"
)
return UAST.Index(), None

elif is_bool(typ_node):
elif _is_bool(typ_node):
if mem_node is not None:
self.err(
node, "size types should not be annotated with " "memory locations"
)
return UAST.Bool(), None

elif is_stride(typ_node):
elif _is_stride(typ_node):
if mem_node is not None:
self.err(
node,
Expand All @@ -403,17 +415,6 @@ def parse_alloc_typmem(self, node):
typ = self.parse_num_type(node)
return typ, mem

_prim_types = {
"R": UAST.Num(),
"f16": UAST.F16(),
"f32": UAST.F32(),
"f64": UAST.F64(),
"i8": UAST.INT8(),
"ui8": UAST.UINT8(),
"ui16": UAST.UINT16(),
"i32": UAST.INT32(),
}

def parse_num_type(self, node, is_arg=False):
if isinstance(node, pyast.Subscript):
if isinstance(node.value, pyast.List):
Expand All @@ -431,23 +432,17 @@ def parse_num_type(self, node, is_arg=False):
)

base = node.value.elts[0]
if (
not isinstance(base, pyast.Name)
or base.id not in Parser._prim_types
):
if not isinstance(base, pyast.Name) or base.id not in _prim_types:
self.err(
node,
"expected window type to be of "
"the form '[R][...]', '[f32][...]', etc.",
)

typ = Parser._prim_types[base.id]
typ = _prim_types[base.id]
is_window = True
elif (
isinstance(node.value, pyast.Name)
and node.value.id in Parser._prim_types
):
typ = Parser._prim_types[node.value.id]
elif isinstance(node.value, pyast.Name) and node.value.id in _prim_types:
typ = _prim_types[node.value.id]
is_window = False
else:
self.err(
Expand Down Expand Up @@ -481,8 +476,14 @@ def parse_num_type(self, node, is_arg=False):

return typ

elif isinstance(node, pyast.Name) and node.id in Parser._prim_types:
return Parser._prim_types[node.id]
elif isinstance(node, pyast.Name) and node.id in _prim_types:
return _prim_types[node.id]
elif isinstance(node, pyast.Name) and (
_is_size(node) or _is_stride(node) or _is_index(node) or _is_bool(node)
):
raise ParseError(
node, f"Cannot allocate an intermediate value of type {node.id}"
)
else:
self.err(node, "unrecognized type: " + pyast.dump(node))

Expand Down
12 changes: 12 additions & 0 deletions tests/test_typecheck.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,6 +4,7 @@

from exo import proc, config
from exo.libs.memories import GEMM_SCRATCH
from exo.pyparser import ParseError


# --- Typechecking tests ---
Expand All @@ -18,6 +19,17 @@ class ConfigLoad:
return ConfigLoad


def test_size0():
with pytest.raises(
ParseError, match="Cannot allocate an intermediate value of type"
):

@proc
def foo(x: size):
size: size
pass


def test_stride1():
ConfigLoad = new_config_ld()

Expand Down
Loading