wenhu1024 64d0deb8ce [VENTUS][fix] Correct edge-case logic for integer to float rounding (#190) ... The conversion script previously applied a complex saturation check to all integer-to-float conversions using the _rtz and _rtn rounding modes. This logic is only necessary for `int` and `uint` to `float` conversions, where precision loss near the maximum value can occur. It was incorrectly applied to other types like `long` or `char`, and to conversions targeting `double`. This commit refines the conditional logic to ensure this specific check is now applied only to the four intended cases: - int -> float (_rtz/_rtn) - uint -> float (_rtz/_rtn)		2025-07-25 16:31:43 +08:00
..
include	[VENTUS][fix] Add mul_hi function	2024-01-22 16:55:53 +08:00
lib	[VENTUS][fix] Correct edge-case logic for integer to float rounding (#190)	2025-07-25 16:31:43 +08:00