Implement statfs with synthetic values

[dannycjones]

Description of change

Draft prototype change for now. May not be merged.

Some applications do not work as expected when a file system reports available space as zero. Some applications may check for available space, and fail if that's a small number (like zero). I believe even df -h will not report the file system in the list by default for some reason - and this may fix it (TBC).

Next steps:

• Are these the right fields to change? What use cases are we solving?
• Confirm that we do actually want to implement statfs by default, diverging from the values provided by Fuser.
• Converge on an agreed value for these fields.
  • u64 max may have some unexpected consequences - we have no idea. Half of u64 is probably large enough.
  • u32 max is too small - that would only be a bit over 4TiB using default block size of 1024.
  • In 2021, AWS shared that there are over 100 trillion objects in Amazon S3. Maybe we could take inspiration from there. https://aws.amazon.com/blogs/aws/amazon-s3s-15th-birthday-it-is-still-day-1-after-5475-days-100-trillion-objects/
• Merge.

Relevant issues: #710

Does this change impact existing behavior?

Mountpoint will now report a non-zero number of available blocks, free blocks, and free files. Before, it reported zero for all three values.

---

By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license and I agree to the terms of the Developer Certificate of Origin (DCO).

[passaro]

How do we think this should show up in df? This is what I now get on AL2:

$ df -ha
Filesystem      Size  Used Avail Use% Mounted on
[...]
mountpoint-s3      0 -4.0Z  4.0Z    - /path/to/mnt

While before it was:

mountpoint-s3      0     0     0    - /path/to/mnt

Note that in both cases I needed the --all flag. Did you expect that to change?

[dannycjones]

How do we think this should show up in df? This is what I now get on AL2:

$ df -ha
Filesystem      Size  Used Avail Use% Mounted on
[...]
mountpoint-s3      0 -4.0Z  4.0Z    - /path/to/mnt

While before it was:

mountpoint-s3      0     0     0    - /path/to/mnt

Note that in both cases I needed the --all flag. Did you expect that to change?

I was hoping it would change based only on previous discussions a while back. Looks like it doesn't.

What's also unexpected there is that we have a negative used value which I imagine we'd want to avoid.

[dannycjones]

Closing this PR for now since I have no intention at the moment of working on it further. However, hopefully this is a good starting point if we do want to move forward with it.