[ZhishengZeng]

What Changed

• Enforce each routing layer's LEF AREA rule when selecting generated power-via geometry.
• Grow only deficient enclosure dimensions, aligned to the manufacturing grid and bounded by the original via footprint. Try both axes; preserve already-compliant vias and reject infeasible candidates instead of widening rails or stripes.
• Keep generated IDB shapes, cached via stacks and DEF export consistent.
• Add regressions for narrow pads, rotated footprints, manufacturing grids, two-axis growth, infeasible footprints, unchanged wide vias, complete stacks and DEF round trips.
• Follow-up to fix(ifp): preserve rail widths and handle narrow endcap rows #340; this PR contains only commit 257a50a3c.

Scope

• C++ core/database/EDA integration
• Python API or native bindings
• Geometry, placement, timing, DRC, or evaluation behavior
• Build/package - CMake, wheel, manylinux, or dependencies
• CI/release - GitHub Actions, version checks, or release automation
• Tests/docs only

Runtime And Packaging Impact

• No runtime or packaging impact
• C++ runtime behavior or native ABI changed
• Python API or extension module behavior changed
• Wheel contents or supported platform changed
• CMake, compiler, dependency, or toolchain requirement changed
• Release artifact or version metadata changed

Notes:

• Changes are confined to iFP. No Tcl/Python API or configuration changes, dependency changes, DRC rule relaxation or process/case-specific production constants.
• An undersized candidate that cannot satisfy AREA inside its footprint is rejected. Existing connectivity validation remains active.
• Rail/stripe widths remain unchanged. Existing longitudinal stripe extension may cover the enlarged legal pad.

Validation

• C++ formatting: automated formatter not run; reviewed scoped changes and ran git diff --check.
• Native build: cmake --build build --target test_ifp_pdn_generator test_ifp_phy_placer test_ifp_config ecc_bin ecc_py --parallel 12.
• Wheel build: not run.
• Focused Python API integration: python -m pytest test/operations/test_python_operations.py::test_floorplan -q passed using the rebuilt extension.
• Focused native tests: ctest --test-dir build -R '^test_ifp_(config|pdn_generator|phy_placer)$' --output-on-failure: 3/3 passed.
• Actual floorplan-stage iDRC before/after using the same checker and technology:
  • William_MCU utilization 0.50: 592 minimum-area violations -> 0 total DRC.
  • William_MCU utilization 0.45: 624 minimum-area violations -> 0 total DRC.
  • Ppu, Multiplier and ecos_pad_ring: 0 -> 0 total DRC.
• Original four fixtures completed DEF write/read/write. ROW, DIEAREA, COMPONENTS, PINS and NETS remained unchanged; rail widths stayed 160 DBU; PG types and name separation were preserved; no ROUTEHALO output.
• All four retain byte-identical power.ploc files and zero floating-rail fragments in the supplied independent coverage screen.
• William_MCU's corrected intermediate-layer pads are 370 x 142 DBU instead of 370 x 100. One edge stripe extends longitudinally by 26 DBU through existing coverage logic; it is not widened.
• git diff --check.

Skipped checks and remaining risk:

• Full rerouting and external signoff DRC/LVS/EMIR were not rerun. The results above are floorplan-stage iDRC, focused native/Python tests and connectivity/coverage regression, not a full signoff claim.
• Distributable wheel and the complete Python test suite were not run.
• Private designs, PDK files, generated artifacts and agent state are not included.

Checklist

• I kept the change scoped to ecc-tools/iFP.
• Behavior changes and validation limits are documented above; no user-facing option changes.
• No dependency/version metadata changes are required.
• No new CMake, compiler or platform requirements.
• I did not include local caches, virtual environments, or generated build outputs.
• I explained skipped validation and remaining risk.