NXP backend: Add pre-processing pass to fuse Lienar + Add#14112

Merged

robert-kalmar merged 2 commits intopytorch:mainfrom

nxp-upstream:upstream/main-nxp/EIEX-471-move-linear-add-fusing-to-executorch

Sep 17, 2025

Commits on Sep 17, 2025

NXP backend: Add pre-processing pass on the aten dialect level to fuse `linear` and `add` nodes.
MartinPavella
committed
NXP backend: Remove IR optimization to fuse FullyConnected and Add operators.
MartinPavella
committed