feat: add volatile tracked query caching

[MichaReiser]

Add an opt-in volatile capacity for tracked queries so large, short-lived values can be collected within a revision.

Volatile queries reuse SIEVE victim selection from #1226 but reclaim selected memo values immediately; owned handles keep concurrent readers safe while metadata remains available for validation and tracked-ID reuse.

Testing: Passed Clippy, workspace, persistence, manual-registration, Shuttle, documentation, and targeted Miri checks.

[netlify]

✅ Deploy Preview for salsa-rs ready!

Name	Link
🔨 Latest commit	5cf55fa
🔍 Latest deploy log	https://app.netlify.com/projects/salsa-rs/deploys/6a496ebba6e585000877ddce
😎 Deploy Preview	https://deploy-preview-1135--salsa-rs.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify project configuration.

[codspeed-hq]

Merging this PR will degrade performance by 82.16%

⚡ 2 improved benchmarks
❌ 12 regressed benchmarks
✅ 60 untouched benchmarks
🆕 13 new benchmarks
⏩ 12 skipped benchmarks¹

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	interned_field_read_many	211.2 µs	837,242.9 µs	-99.97%
❌	Simulation	input_field_read_many	260.1 µs	706,622.7 µs	-99.96%
❌	Simulation	tracked_field_read_many	271 µs	135,703.6 µs	-99.8%
❌	Simulation	untracked_field_read_many	220.7 µs	40,200.6 µs	-99.45%
❌	Simulation	fill_and_evict[Lru]	443.5 µs	621.1 µs	-28.61%
❌	Memory	fill_and_evict[NoEviction]	23.4 KB	25.9 KB	-9.64%
❌	Simulation	fill_and_evict[NoEviction]	379.1 µs	413.4 µs	-8.29%
❌	Simulation	converge_diverge_nested	130.8 µs	142.3 µs	-8.08%
❌	Memory	fill_and_evict[Lru]	35.5 KB	38 KB	-6.58%
❌	Simulation	converge_diverge	152.6 µs	162.2 µs	-5.92%
❌	WallTime	project_check_then_incremental[Lru]	9.8 ms	10.4 ms	-5.51%
❌	WallTime	parallel_fast_path[Lru]	3.8 ms	3.9 ms	-4.35%
⚡	Simulation	cold_execute	266.5 ms	5.4 ms	×49
⚡	Simulation	deep_verify_then_execute	4.9 ms	3.2 ms	+53.9%
🆕	WallTime	one_hit_wonders[Sieve]	N/A	1 ms	N/A
🆕	WallTime	parallel_fast_path[Sieve]	N/A	409 µs	N/A
🆕	WallTime	phase_change[Sieve]	N/A	353.7 µs	N/A
🆕	WallTime	project_check_then_incremental[Sieve]	N/A	6.9 ms	N/A
🆕	WallTime	scan_resistance[Sieve]	N/A	1.2 ms	N/A
🆕	Memory	disabled_eviction[Sieve]	N/A	0 B	N/A
...	...	...	...	...	...

ℹ️ Only the first 20 benchmarks are displayed. Go to the app to view all benchmarks.

Tip

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

---

_{Comparing MichaReiser:volatile-query-cache (5cf55fa) with master (caa352f)}

Footnotes

1. 12 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. ↩