dd: simplify signal handling by removing Alarm timer thread

[ChrisDryden]

We have been experiencing intermittent failures in the dd unit tests, that are caused by the utility exiting with the SIGPIPE signal at the end of running and not returning an exit code. This is caused by the fact that this utility used pipes internally to handle signals and because it assumes that all rust programs ignore SIGPIPE.

When investigating the implementation of the signal handling, it appeared to me that the implementation was over-complicating things by having two separate threads involving pipes and atomic locks for timers that it could be very simplified by using the approach of just using a signal handler.

Fixes #10442
Fixes #10455
Fixes #10766

[github-actions]

GNU testsuite comparison:

GNU test failed: tests/factor/t30. tests/factor/t30 is passing on 'main'. Maybe you have to rebase?
GNU test failed: tests/factor/t31. tests/factor/t31 is passing on 'main'. Maybe you have to rebase?
GNU test failed: tests/factor/t35. tests/factor/t35 is passing on 'main'. Maybe you have to rebase?
Skipping an intermittent issue tests/misc/usage_vs_getopt (passes in this run but fails in the 'main' branch)

[codspeed-hq]

CodSpeed Performance Report

Merging this PR will improve performance by 7%

_{Comparing ChrisDryden:dd-simplify-signal-handling (54dfa0b) with main (da7ad84)}

Summary

⚡ 8 improved benchmarks
✅ 276 untouched benchmarks
⏩ 38 skipped benchmarks¹

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
⚡	Memory	dd_copy_default	105.9 KB	99 KB	+7%
⚡	Memory	dd_copy_separate_blocks	185.3 KB	177.3 KB	+4.53%
⚡	Memory	dd_copy_with_skip	109.6 KB	102.7 KB	+6.75%
⚡	Memory	dd_copy_8k_blocks	113.3 KB	106.3 KB	+6.52%
⚡	Memory	dd_copy_partial	109.3 KB	102.4 KB	+6.77%
⚡	Memory	dd_copy_64k_blocks	169.3 KB	162.3 KB	+4.27%
⚡	Memory	dd_copy_with_seek	109.6 KB	102.7 KB	+6.75%
⚡	Memory	dd_copy_4k_blocks	109.3 KB	102.3 KB	+6.77%

Footnotes

1. 38 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. ↩

[ChrisDryden]

Darn, the performance numbers are unfortunate, I'll have to see where this is coming from

[ChrisDryden]

Theres two changes here, one to remove the alarm thread and one to remove the pipe based signal handler, I can simplify this to just remove the pipe based signal handler.

[github-actions]

GNU testsuite comparison:

Skip an intermittent issue tests/misc/usage_vs_getopt (fails in this run but passes in the 'main' branch)
Skip an intermittent issue tests/shuf/shuf-reservoir (fails in this run but passes in the 'main' branch)
Skip an intermittent issue tests/sort/sort-stale-thread-mem (fails in this run but passes in the 'main' branch)

[ChrisDryden]

Sweet lower memory and no performance regressions.

[Ecordonnier]

While reviewing I've noticed we're missing macOS SIGINFO handling, so I've created #10776 as follow-up.

[Ecordonnier]

we can also remove it from deny.toml AFAIU, since there are no version conflicts any more

[github-actions]

GNU testsuite comparison:

Skip an intermittent issue tests/tail/inotify-dir-recreate (fails in this run but passes in the 'main' branch)