Module int_lane 

The certified-EXACT integer lane — small-range integer dataflow with the ADC as restorer.

Integer programs extract as integer ALU cells (Arith), and their constants are VALUES (Src::Const(7) is seven, not a bit pattern) — the mirror image of the float lane’s conventions. The tile vocabulary is unchanged: integer add/sub is the weighted Kirchhoff sum, integer-constant multiply a programmed gain, variable×variable the four-quadrant multiplier. What is NEW is the claim: because every weight and constant is an integer, the region is GRID-PRESERVING, and when the compiler’s error budget proves ε < ½ grid step (super::exactness_gate) the rounded readout is EXACTLY the VM’s integer — a stronger statement than the float lane’s ε-≡, purchased with the grid’s slack. Division, remainder, and every bitwise op refuse: quotient grids are unforgiving and bit patterns are not voltages.

Functions

analog_int_netlist_from_source

Lower an INTEGER function to an analog tile netlist realizable on target. The exactness claim itself is made downstream by super::exactness_gate over the scheduled budget.