fix(trap): report signals ignored on shell entry in trap -p

[lu-zero]

Summary

Bash reports signals that were ignored (SIG_IGN disposition) when the shell started — e.g. trap -- '' SIGRTMIN — via trap -p. Brush currently only prints explicitly registered handlers, so environments that ignore a signal on entry (like the GitHub Actions runner ignoring SIGRTMIN) produce a trap -p diff against bash.

This PR makes brush capture the ignored-on-entry signals at shell creation and report them in trap -p, matching bash.

Changes

• brush-core/src/sys/unix/signal.rs: new ignored_signals_on_entry() queries each signal's disposition via sigaction, including real-time signals (SIGRTMIN..SIGRTMAX) on Linux/Android.
  • SIGPIPE is excluded: the Rust runtime always sets it to SIG_IGN at startup for every Rust program, which does not reflect the parent's disposition, while bash (a C program) starts with SIG_DFL in comparable spawn scenarios.
• brush-core/src/sys/stubs/signal.rs: stub returns an empty set.
• brush-core/src/traps.rs: TrapHandlerConfig stores the ignored-on-entry signal numbers; new signal_name_for_number() names signals the way bash does, including real-time naming (SIGRTMIN, SIGRTMIN+1, SIGRTMAX-1, …).
• brush-core/src/shell.rs: captures ignored-on-entry signals in Shell::new.
• brush-builtins/src/trap.rs: trap -p and trap -p <sig> report ignored-on-entry signals as trap -- '' <SIG>.

Validation

• Full compat suite run under a parent that ignores SIGRTMIN (replicating the CI environment): 2174 test cases, 0 failed (the previously-failing trap -p - with no args shows all traps case now passes).
• cargo clippy and cargo fmt --check clean; brush-core unit tests pass.

Fixes the CI failure seen on the OS-target test jobs in e.g. #1267.

[github-actions]

Public API changes for crate: brush-core

Added items

+pub fn brush_core::sys::signal::ignored_signals_on_entry() -> alloc::vec::Vec<i32>
+pub fn brush_core::traps::TrapHandlerConfig::is_signal_ignored_on_entry(&self, i32) -> bool
+pub fn brush_core::traps::TrapHandlerConfig::iter_signals_ignored_on_entry(&self) -> impl core::iter::traits::iterator::Iterator<Item = i32> + '_
+pub fn brush_core::traps::TrapHandlerConfig::set_signals_ignored_on_entry(&mut self, alloc::vec::Vec<i32>)
+pub fn brush_core::traps::signal_name_for_number(i32) -> alloc::string::String

Performance Benchmark Report

Benchmark name	Baseline (μs)	Test/PR (μs)	Delta (μs)	Delta %
clone_shell_object	18.59 μs	18.53 μs	-0.07 μs	⚪ Unchanged
eval_arithmetic	0.13 μs	0.13 μs	0.00 μs	⚪ Unchanged
expand_one_string	1.71 μs	1.65 μs	-0.06 μs	⚪ Unchanged
for_loop	29.30 μs	28.92 μs	-0.39 μs	🟢 -1.32%
full_peg_complex	54.78 μs	54.70 μs	-0.08 μs	⚪ Unchanged
full_peg_for_loop	6.28 μs	6.37 μs	0.08 μs	⚪ Unchanged
full_peg_nested_expansions	15.60 μs	15.76 μs	0.16 μs	⚪ Unchanged
full_peg_pipeline	4.14 μs	4.24 μs	0.09 μs	🟠 +2.24%
full_peg_simple	1.69 μs	1.73 μs	0.04 μs	🟠 +2.49%
function_call	3.71 μs	3.63 μs	-0.08 μs	⚪ Unchanged
instantiate_shell	54.38 μs	63.57 μs	9.19 μs	🟠 +16.90%
instantiate_shell_with_init_scripts	25124.78 μs	24593.48 μs	-531.30 μs	⚪ Unchanged
parse_peg_bash_completion	2224.14 μs	2245.13 μs	20.99 μs	🟠 +0.94%
parse_peg_complex	19.08 μs	19.15 μs	0.07 μs	⚪ Unchanged
parse_peg_for_loop	1.89 μs	1.89 μs	0.00 μs	⚪ Unchanged
parse_peg_pipeline	2.06 μs	1.94 μs	-0.12 μs	🟢 -5.83%
parse_peg_simple	1.05 μs	1.05 μs	0.00 μs	⚪ Unchanged
run_echo_builtin_command	13.16 μs	13.14 μs	-0.02 μs	⚪ Unchanged
tokenize_sample_script	3.44 μs	3.50 μs	0.07 μs	🟠 +1.95%

Code Coverage Report: Only Changed Files listed

Package	Base Coverage	New Coverage	Difference
brush-builtins/src/trap.rs	🟢 98.51%	🟢 82.76%	🔴 -15.75%
brush-core/src/results.rs	🟢 83.33%	🟢 80.16%	🔴 -3.17%
brush-core/src/shell.rs	🟢 94.85%	🟢 94.92%	🟢 0.07%
brush-core/src/sys/unix/signal.rs	🟠 60%	🟠 71.43%	🟢 11.43%
brush-core/src/traps.rs	🟢 86.87%	🟢 91.1%	🟢 4.23%
Overall Coverage	🟢 76.18%	🟢 76.19%	🟢 0.01%

Minimum allowed coverage is 70%, this run produced 76.19%
Maximum allowed coverage difference is -5%, this run produced 0.01%

Test Summary: bash-completion test suite

Outcome	Count	Percentage
✅ Pass	1582	75.01
❗️ Error	17	0.81
❌ Fail	156	7.40
⏩ Skip	339	16.07
❎ Expected Fail	13	0.62
✔️ Unexpected Pass	2	0.09
📊 Total	2109	100.00