Haiku uses a i64 for pointers within statvfs. I was working on patches to support Haiku and ran across this:
fn free_blocks(&self) -> u64 {
#[cfg(target_pointer_width = "64")]
return self.f_bfree;
#[cfg(not(target_pointer_width = "64"))]
return self.f_bfree.into();
}This works for u64 vs not u64, however haiku uses i64 which means this code fails to compile.
I worked up a fix to address this... however it is pretty ugly:
fn total_blocks(&self) -> u64 {
#[cfg(all(
target_pointer_width = "64",
not(target_os = "haiku")
))]
return self.f_blocks;
#[cfg(any(
not(target_pointer_width = "64"),
target_os = "haiku",
))]
return self.f_blocks.into();
}I feel like these could all be condensed into just return self.f_blocks.into(); which would cover all cases, but i'm also assuming the width check is for a reason?
Haiku uses a i64 for pointers within statvfs. I was working on patches to support Haiku and ran across this:
This works for u64 vs not u64, however haiku uses i64 which means this code fails to compile.
I worked up a fix to address this... however it is pretty ugly:
I feel like these could all be condensed into just
return self.f_blocks.into();which would cover all cases, but i'm also assuming the width check is for a reason?