bench: use seeded RNG for glass_pumpkin prime generation benchmarks

[sylvestre]

Use glass_pumpkin's from_rng() API with a seeded ChaCha8Rng instead of new() which uses OsRng internally, making these benchmarks deterministic. num-primes doesn't expose an RNG parameter so it remains non-deterministic.

[codspeed-hq]

Merging this PR will degrade performance by 92.34%

⚡ 2 improved benchmarks
❌ 2 regressed benchmarks
✅ 25 untouched benchmarks

⚠️ Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
⚡	Simulation	glass_pumpkin	3,163.2 µs	984 µs	×3.2
⚡	Simulation	glass_pumpkin	33.9 ms	10.5 ms	×3.2
❌	Simulation	num-primes	22 ms	28.9 ms	-23.71%
❌	Simulation	num-primes	759.5 ms	9,920 ms	-92.34%

---

_{Comparing sylvestre:rand-2 (35a4b6c) with main (75e7ea8)}

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 84.66%. Comparing base (3f5db7d) to head (35a4b6c).
⚠️ Report is 18 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #57      +/-   ##
==========================================
+ Coverage   78.64%   84.66%   +6.01%     
==========================================
  Files          13       14       +1     
  Lines        2585     3403     +818     
  Branches      230      238       +8     
==========================================
+ Hits         2033     2881     +848     
+ Misses        552      521      -31     
- Partials        0        1       +1     

Flag	Coverage Δ
macos_latest	84.57% <ø> (+6.93%)	⬆️
ubuntu_latest	84.21% <ø> (+6.03%)	⬆️
windows_latest	17.86% <ø> (+3.16%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.