dd: don't panic on huge count/skip/seek operands

[mistaste]

Summary

Fixes #12843, #12845, #12846, #12848.

A very large count=, skip=, seek=, iseek= or oseek= operand overflowed
an internal blocks * block_size multiplication and aborted the process:

$ dd if=a of=b count=177817277272177278
thread 'main' panicked at src/uu/dd/src/dd.rs: attempt to multiply with overflow
$ dd if=a of=b iseek=166727721627616621762772
thread 'main' panicked at src/uu/dd/src/dd.rs: attempt to multiply with overflow

Change

• Num::to_bytes now uses saturating_mul, so an out-of-range block count
  saturates to u64::MAX and is caught by the existing > i64::MAX (intmax_t)
  limit checks for skip/seek instead of panicking.
• The same i64::MAX limit is now also enforced for count, which previously
  had no check and overflowed later in calc_loop_bsize.

These operands now report a clean error, matching GNU:

$ dd count=177817277272177278
dd: invalid number: '...': Value too large for defined data type

Checks

• New regression test test_large_numeric_operands_do_not_panic (covers count/skip/seek/iseek/oseek); existing dd tests still pass.
• cargo fmt clean, cargo clippy -p uu_dd --all-targets clean.

[github-actions]

GNU testsuite comparison:

Skip an intermittent issue tests/cut/bounded-memory (fails in this run but passes in the 'main' branch)
Skip an intermittent issue tests/date/date-locale-hour (fails in this run but passes in the 'main' branch)
Skipping an intermittent issue tests/pr/bounded-memory (passes in this run but fails in the 'main' branch)
Note: The gnu test tests/env/env-signal-handler was skipped on 'main' but is now failing.