fsext: Cleanup; cut down on unnecessary macros

[kallisti5]

• Some platforms like Haiku offer a i64 for statvfs pointers.
• The pointer-width checks don't catch i64 vs u64
• Instead of the duplicated code and excessive platform checks, prefer 'as u64' which catches all situations

[kallisti5]

This one is based on the conversation in #14607 Cuts way, way, way down on duplicate code while preserving platform compatibility. (hopefully!)

• Unit tests run on Linux ok and code builds. Unable to test on a lot of the "random less known architectures" however.

[xtqqczze]

@kallisti5 sorry, I already had fully tested changes to get fsext to compile on Haiku and NetBSD: #14627

[kallisti5]

In the context of fsext... I feel like this one is a little cleaner 😆 It should accomplish the same thing and work across more platforms more universally.

[xtqqczze]

Sorry, but I think we lose some type-safety with these simplifications. I assume the original code was structured that way for a reason.

[kallisti5]

Rust is type-safe at compile though, so that doesn't really make sense.
Overall, the look of this code is "everyone was too afraid to mess with other platforms, so they stacked macros"

I'm 100% sure that there was a reason originally for this design choice, but everyone just stacked tech-debt on-top as they added additional platforms to support.

The rest of the Haiku patches look awesome, but I feel like we have a good opportunity here to cleanup some major tech-debt in the process.

@sylvestre absolutely feel free to disagree with me here though. I'm a newbie making contributions to coreutils

[github-actions]

GNU testsuite comparison:

GNU test failed: tests/csplit/csplit. tests/csplit/csplit is passing on 'main'. Maybe you have to rebase?
GNU test failed: tests/df/df-output. tests/df/df-output is passing on 'main'. Maybe you have to rebase?
GNU test failed: tests/du/max-depth. tests/du/max-depth is passing on 'main'. Maybe you have to rebase?
GNU test failed: tests/fold/fold. tests/fold/fold is passing on 'main'. Maybe you have to rebase?
GNU test failed: tests/head/quote-headers. tests/head/quote-headers is passing on 'main'. Maybe you have to rebase?
GNU test failed: tests/ls/ls-time. tests/ls/ls-time is passing on 'main'. Maybe you have to rebase?
GNU test failed: tests/ls/m-option. tests/ls/m-option is passing on 'main'. Maybe you have to rebase?
GNU test failed: tests/misc/just-created. tests/misc/just-created is passing on 'main'. Maybe you have to rebase?
GNU test failed: tests/mv/part-hardlink-symlink. tests/mv/part-hardlink-symlink is passing on 'main'. Maybe you have to rebase?
GNU test failed: tests/printf/printf. tests/printf/printf is passing on 'main'. Maybe you have to rebase?
GNU test failed: tests/ptx/word-regex-loop. tests/ptx/word-regex-loop is passing on 'main'. Maybe you have to rebase?
GNU test failed: tests/seq/seq-precision. tests/seq/seq-precision is passing on 'main'. Maybe you have to rebase?
GNU test failed: tests/shuf/shuf. tests/shuf/shuf is passing on 'main'. Maybe you have to rebase?
GNU test failed: tests/sort/sort-buffer-size. tests/sort/sort-buffer-size is passing on 'main'. Maybe you have to rebase?
GNU test failed: tests/split/split-io-err. tests/split/split-io-err is passing on 'main'. Maybe you have to rebase?
GNU test failed: tests/stat/stat-fmt. tests/stat/stat-fmt is passing on 'main'. Maybe you have to rebase?
GNU test failed: tests/stat/stat-nanoseconds. tests/stat/stat-nanoseconds is passing on 'main'. Maybe you have to rebase?
GNU test failed: tests/tail/quote-headers. tests/tail/quote-headers is passing on 'main'. Maybe you have to rebase?
GNU test failed: tests/test/test. tests/test/test is passing on 'main'. Maybe you have to rebase?
Skip an intermittent issue tests/tail/pipe-f (fails in this run but passes in the 'main' branch)
Congrats! The gnu test tests/df/skip-rootfs is no longer failing!
Congrats! The gnu test tests/id/smack is no longer failing!
Congrats! The gnu test tests/mkdir/smack-root is no longer failing!
Note: The gnu test tests/mv/i-3 is now being skipped but was previously passing.
Note: The gnu test tests/stty/bad-speed is now being skipped but was previously passing.
Note: The gnu test tests/stty/stty is now being skipped but was previously passing.
Note: The gnu test tests/stty/stty-invalid is now being skipped but was previously passing.
Note: The gnu test tests/stty/stty-pairs is now being skipped but was previously passing.
Note: The gnu test tests/stty/stty-row-col is now being skipped but was previously passing.
Note: The gnu test tests/misc/getopt_vs_usage was skipped on 'main' but is now failing.
Note: The gnu test tests/misc/usage_vs_refs was skipped on 'main' but is now failing.
Note: The gnu test tests/rm/rm-readdir-fail was skipped on 'main' but is now failing.

[xtqqczze]

For example, this case would previously have triggered a panic on x86_64-unknown-netbsd, but no longer does after this change:

$ RUSTC_BOOTSTRAP=1 cargo clippy -q -Zbuild-std -p uucore --features fsext --all-targets --target x86_64-unknown-netbsd -- -D clippy::cast_possible_wrap
error: casting `u64` to `i64` may wrap around the value
   --> src/uucore/src/lib/features/fsext/mod.rs:485:9
    |
485 |         self.f_bsize as i64
    |         ^^^^^^^^^^^^^^^^^^^ help: if this is intentional, use `cast_signed()` instead: `self.f_bsize.cast_signed()`
    |
    = help: for further information visit https://rust-lang.github.io/rust-clippy/rust-1.98.0/index.html#cast_possible_wrap

[xtqqczze]

Sorry, but I think we lose some type-safety with these simplifications. I assume the original code was structured that way for a reason.

@kallisti5 Thank you for pushing back on this. I considered your comments and made some simplifications using try_into().unwrap() and into() for the fs feature in 17f1a5d.

[github-actions]

Binary size comparison:

Individual binary size comparison VS main (threshold: >=5% AND >=4 KB).

Total size of compared binaries: 153.03 MB (-2.66 MB, -1.71%)

Significant per-binary changes:
  [        2.32 MB ->    1.12 MB  (-1.20 MB, -51.76%)
  test     2.32 MB ->    1.12 MB  (-1.20 MB, -51.76%)

[codspeed]

Merging this PR will regress 2 benchmarks

⚠️ Different runtime environments detected

Some benchmarks with significant performance changes were compared across different runtime environments,
which may affect the accuracy of the results.

Open the report in CodSpeed to investigate

⚡ 11 improved benchmarks
❌ 2 regressed benchmarks
✅ 360 untouched benchmarks
⏩ 50 skipped benchmarks¹

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	sort_dictionary_order[500000]	1.9 s	2.1 s	-6.43%
❌	Simulation	ls_recursive_deep_tree[(200, 2)]	3.2 ms	3.3 ms	-4.09%
⚡	Simulation	complex_relative_date	397.5 µs	287 µs	+38.49%
⚡	Simulation	check_sorted_utf8_locale	492.2 ms	422.9 ms	+16.4%
⚡	Simulation	sort_numeric_utf8_locale	41.2 ms	36.5 ms	+12.61%
⚡	Simulation	merge_single_file_utf8_locale	133.3 ms	119.4 ms	+11.66%
⚡	Simulation	merge_pre_sorted_files_utf8_locale	251.9 ms	238 ms	+5.88%
⚡	Simulation	merge_pre_sorted_files	252.8 ms	238.9 ms	+5.83%
⚡	Simulation	sort_case_sensitive[500000]	339.5 ms	325.7 ms	+4.22%
⚡	Simulation	sort_german_de_locale	297.9 ms	285.8 ms	+4.21%
⚡	Simulation	du_wide_tree[(5000, 500)]	18.7 ms	18.1 ms	+3.57%
⚡	Simulation	du_all_wide_tree[(5000, 500)]	38.2 ms	36.9 ms	+3.42%
⚡	Simulation	sort_spill_to_tmp_files_utf8_locale	340.2 ms	329.1 ms	+3.39%

Tip

Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing kallisti5:fxext-cleanup (8a7f8eb) with main (529f011)}

Footnotes

1. 50 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩