mirror of
https://github.com/mirror/tinycc.git
synced 2026-08-22 11:23:27 +08:00
when x86-64 support was added, "for long double, we use x87 FPU". And indeed, tests show that Intel's extended precision is used, not double precision.