Ignore the FLATC_EXECUTABLE env var when empty (#3612)

dbort · facebook-github-bot · commit 2b1ed26164ea · 2024-05-15T14:25:08.000-07:00
Summary: We've seen some situations fail because flatc_path is an empty string. This seems like a likely culprit, but doesn't hurt either way. Pull Request resolved: #3612 Test Plan: Ran the following script with `(. /tmp/run.sh)`, which successfully created `/tmp/et-test/executorch/add.pte`. `/tmp/run.sh`: ``` set -euxo pipefail conda remove -yn et-test --all conda create -yn et-test python=3.10 conda activate et-test mkdir -p /tmp/et-test cd /tmp/et-test # Get this PR git clone -b flatc-empty-path https://github.com/dbort/executorch-1 cd executorch git submodule sync git submodule update --init ./install_requirements.sh cat > /tmp/export_add.py << HERE import torch from torch.export import export from executorch.exir import to_edge # Start with a PyTorch model that adds two input tensors (matrices) class Add(torch.nn.Module): def __init__(self): super(Add, self).__init__() def forward(self, x: torch.Tensor, y: torch.Tensor): return x + y # 1. torch.export: Defines the program with the ATen operator set. aten_dialect = export(Add(), (torch.ones(1), torch.ones(1))) # 2. to_edge: Make optimizations for Edge devices edge_program = to_edge(aten_dialect) # 3. to_executorch: Convert the graph to an ExecuTorch program executorch_program = edge_program.to_executorch() # 4. Save the compiled .pte program with open("add.pte", "wb") as file: file.write(executorch_program.buffer) HERE python /tmp/export_add.py echo "Created" *.pte ``` Reviewed By: lucylq Differential Revision: D57390281 Pulled By: dbort fbshipit-source-id: 91c507764ef461f1988d7f3c45579120d9f6ed96
diff --git a/exir/_serialize/_flatbuffer.py b/exir/_serialize/_flatbuffer.py
@@ -201,7 +201,9 @@ def _run_flatc(args: Sequence[str]) -> None:
             subprocess.run([flatc_path] + list(args), check=True)
     else:
         # Expect the `flatc` tool to be on the system path or set as an env var.
-        flatc_path = os.getenv("FLATC_EXECUTABLE", "flatc")
+        flatc_path = os.getenv("FLATC_EXECUTABLE")
+        if not flatc_path:
+            flatc_path = "flatc"
         subprocess.run([flatc_path] + list(args), check=True)
 
 
