Skip to content

Fix noop pass when there are dynamic shapes #17678

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

Open
metascroy wants to merge 2 commits into
base: main
Choose a base branch
from
+9 −3
Open

Fix noop pass when there are dynamic shapes #17678

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
1c84261
up
metascroy Feb 24, 2026
0044a09
Merge branch 'main' into fix-noop-pass
metascroy Feb 26, 2026
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
12 changes: 9 additions & 3 deletions exir/passes/remove_noop_pass.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -77,9 +77,15 @@ def call(self, graph_module: GraphModule) -> PassResult:
continue

if node.target == torch.ops.aten.slice_copy.Tensor:
# Only do this check if all the dims are static.
if all(isinstance(dim, int) for dim in orig_tensor.size()):
if orig_tensor.shape == node.meta["val"].shape:
output_tensor = node.meta["val"]
# Only do this check if all dims are static on both sides.
# The output may contain unbacked SymInts (e.g. from
# data-dependent slicing with .item()), so we must check
# both tensors before comparing shapes.
if all(isinstance(dim, int) for dim in orig_tensor.size()) and all(
isinstance(dim, int) for dim in output_tensor.size()
):
if orig_tensor.shape == output_tensor.shape:
# If the graph is quantized, we must remove the entire pattern consisting of dq->op->q.
# Otherwise, removing only the op will suffice.
if node.args[0].target in _DEQUANT_OPS:
Expand Down
Loading