v3.1: accounts-db: Use SmallRng for LRU eviction sampling (backport of #10640)

[mergify]

Problem

Sampling is the biggest bottleneck in the eviction mechanism, which with the default Rng takes 82% of the whole eviction process and causes the overall eviction process to take over 20s.

Summary of Changes

Optimize it by using SmallRng.

---

☝️ The new profiles are still not ideal - the lock contention, previously taking 11% of time, now is the biggest bottleneck taking 91%. But that's going away in #10641.

---

This is an automatic backport of pull request #10640 done by Mergify.

[mergify]

Cherry-pick of 1ffcdcc has failed:

On branch mergify/bp/v3.1/pr-10640
Your branch is up to date with 'origin/v3.1'.

You are currently cherry-picking commit 1ffcdcc60.
  (fix conflicts and run "git cherry-pick --continue")
  (use "git cherry-pick --skip" to skip this patch)
  (use "git cherry-pick --abort" to cancel the cherry-pick operation)

Unmerged paths:
  (use "git add <file>..." to mark resolution)
	both modified:   accounts-db/src/read_only_accounts_cache.rs

no changes added to commit (use "git add" and/or "git commit -a")

To fix up this pull request, you can check it out locally. See documentation: https://docs.github.com/en/pull-requests/collaborating-with-pull-requests/reviewing-changes-in-pull-requests/checking-out-pull-requests-locally

[t-nelson]

@vadorovsky mind resolving merge conflicts here?

[vadorovsky]

@vadorovsky mind resolving merge conflicts here?

Done. To elaborate on what I had to fix, the ctor for an RNG seeded from getrandom has different names in different rand_core versions.

In rand_core-0.6.4 it's called SeedableRng::from_entropy and looks like:

https://github.com/rust-random/rand_core/blob/89a1336b934c68ddce548127c6f8afd910b35a18/rand_core/src/lib.rs#L410-L418Th

In rand_core-0.9.3 it's called SeedableRng::from_os_rng with the following definition:

https://github.com/rust-random/rand_core/blob/2d84d25c45a68c1424e4fa775d8a527dc16cec65/src/lib.rs#L554-L577

I don't want that info to get lost - is it fine if after approval, before merging the PR, I squash the commits and add the info on the bottom of the commit message?

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 83.2%. Comparing base (01a77c7) to head (cc69c00).

Additional details and impacted files

@@            Coverage Diff            @@
##             v3.1   #10644     +/-   ##
=========================================
- Coverage    83.2%    83.2%   -0.1%     
=========================================
  Files         865      865             
  Lines      376601   376601             
=========================================
- Hits       313609   313550     -59     
- Misses      62992    63051     +59     

🚀 New features to boost your workflow:

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

[steviez]

I don't want that info to get lost - is it fine if after approval, before merging the PR, I squash the commits and add the info on the bottom of the commit message?

I'm going to merge this but will ensure this information is retained; looks like your additional commit had good information that will leave us a nice papertrail