Model x86 80-bit extended `long double` precisely#9023

Open

tautschnig wants to merge 8 commits into

diffblue:developfrom

tautschnig:x86-long-double-extended

Commits on Jun 9, 2026

smt2: flatten FPA-encoded floats to their IEEE bit pattern

tautschnig
and
kiro-agent
committed
ieee_float, solvers: model x86 80-bit extended long double precisely

tautschnig
and
kiro-agent
committed
Tests: x86 80-bit extended long double layout and round-trips

tautschnig
and
kiro-agent
committed
cbmc-library: fix long-double Schraudolph models for x86 80-bit extended

tautschnig
and
kiro-agent
committed
doc: architectural notes on x86 80-bit extended long double

tautschnig
and
kiro-agent
committed
smt2: encode x86 80-bit extended long double as bit-vectors

tautschnig
and
kiro-agent
committed
smt2: encode mixed FPA/bit-vector float typecasts

tautschnig
and
kiro-agent
committed
smt2: register float_bv typecast helper for bit-vector-encoded results

tautschnig
and
kiro-agent
committed