[v0.22.3] chore!: rollback proctree to simple LRU (#4299)

[rscampos]

1. Explain what the PR does

d7083d7 chore!: rollback proctree to simple LRU (#4299)

The expirable LRU cache is not intended to be used in the ProcessTree
type since the TTL would impact its correctness on some scenarios.

Despite TTL can be disabled by setting it to 0, it is not a good idea
due to the potential performance impact of its implementation. See the
comparison between the simple LRU and the expirable LRU below:

| Benchmark                              | Exp LRU |  Simple |    %   |
|                                        |  TTL 0  |   LRU   |  Impr. |
|                                        | (ns/op) | (ns/op) |    *   |
|----------------------------------------|---------|---------|--------|
| GetProcessByHash-Concurrency1          | 4.264   | 6.301   | -47.8% |
| GetProcessByHash-Concurrency2          | 14.91   | 19.85   | -33.2% |
| GetProcessByHash-Concurrency4          | 74.03   | 63.22   | 14.6%  |
| GetProcessByHash-Concurrency8          | 247.2   | 174.7   | 29.3%  |
| GetOrCreateProcessByHash-Concurrency1  | 37.02   | 8.370   | 77.4%  |
| GetOrCreateProcessByHash-Concurrency2  | 80.84   | 23.37   | 71.1%  |
| GetOrCreateProcessByHash-Concurrency4  | 181.6   | 75.67   | 58.4%  |
| GetOrCreateProcessByHash-Concurrency8  | 408.3   | 194.5   | 52.4%  |
| GetThreadByHash-Concurrency1           | 38.13   | 7.675   | 79.9%  |
| GetThreadByHash-Concurrency2           | 79.98   | 22.71   | 71.6%  |
| GetThreadByHash-Concurrency4           | 177.1   | 60.76   | 65.7%  |
| GetThreadByHash-Concurrency8           | 403.2   | 190.0   | 52.9%  |
| GetOrCreateThreadByHash-Concurrency1   | 38.16   | 7.996   | 79.0%  |
| GetOrCreateThreadByHash-Concurrency2   | 79.82   | 23.73   | 70.3%  |
| GetOrCreateThreadByHash-Concurrency4   | 177.1   | 67.13   | 62.1%  |
| GetOrCreateThreadByHash-Concurrency8   | 405.1   | 151.3   | 62.7%  |

Therefore, this commit reverts the ProcessTree to use the simple LRU
cache until a better solution is designed.

* Part of the performance improvement is also due to the removal of the
outer lock done by 44e59d3a9731166b5345a3c9c396544810e46a65.

commit: 34be604 (main), cherry-pick

2. Explain how to test it

3. Other comments

[geyslan]

LGTM