fs: support utimes BigInt

[guybedford]

What is the problem this feature will solve?

Current utimes FS functions have a granularity of microseconds, while the stats BigInt functionality has a granularity of nanoseconds.

What is the feature you are proposing to solve the problem?

Adding support for BigInt to the utimes functions to enable nanosecond granularity for setting timestamps.

What alternatives have you considered?

The underlying reason for this issue is that floating point second values only have 24 significant digits, which does not extend to exact nanoseconds. Having stats support BigInt u64 times, while having utimes not creates the inability to exactly set and read the same level of precision. Relying on a BigInt input as a u64 nanoseconds value seems the most sensible way to resolve this to me.

[github-actions]

There has been no activity on this feature request for 5 months. To help maintain relevant open issues, please add the never-stale Mark issue so that it is never considered stale label or close this issue if it should be closed. If not, the issue will be automatically closed 6 months after the last non-automated comment.
For more information on how the project manages feature requests, please consult the feature request management document.

[github-actions]

There has been no activity on this feature request and it is being closed. If you feel closing this issue is not the right thing to do, please leave a comment.

For more information on how the project manages feature requests, please consult the feature request management document.

[hoodmane]

Seems to be an upstream uv problem. uv's stat returns a time_spec struct with a pair of a long number of tv_secs and a long number of tv_nsecs which get combined into a bigint. On the other hand, uv's utimes family of functions take a double and use a function called uv__fs_to_timeval to convert it to a time_spec. (On the other hand, it looks like windows might not support fractional seconds at all? Looks like it just casts the double to a time_t on windows, throwing away the fractional part.) To fix requires adding utimes functions to uv that take a time_spec instead.