Closed chlongv closed 4 years ago
Thanks for the PR @chlongv ! It'll be great to finally be rid of f128. Curious that the f128 version was buggy... that seems like an issue with c2rust.
The CI error is my fault... I'll get that fixed ASAP.
Ok, should be fixed in https://github.com/samuela/rustybox/commit/a1360f80cfff422cdf64ad3f63156ac6f79ba512
Thanks for the PR @chlongv ! It'll be great to finally be rid of f128. Curious that the f128 version was buggy... that seems like an issue with c2rust.
Have a look at this thread: https://github.com/rust-lang/rust-bindgen/issues/1549
Doing some tests I have noticed the difference: with a binary 0.0 long double (confirmed with the "real" od), rustybox's od gave me nan (not a number).
Remove -t fL support for the od applet. It was anyhow non-functional as the use f128 representation was not compatible with the c representation, resulting in different results between rustybox and the classic od.
This allows to remove the f128 dependency as it is the only use of it and this should help for the cross-compilation that currently fails with f128.