Introduce BOP_CMP for optimized comparison

Prior to this commit the `OPTIMIZED_CMP` macro relied on a method lookup
to determine whether `<=>` was overridden. The result of the lookup was
cached, but only for the duration of the specific method that
initialized the cmp_opt_data cache structure.

With this method lookup, `[x,y].max` is slower than doing `x > y ?
x : y` even though there's an optimized instruction for "new array max".
(John noticed somebody a proposed micro-optimization based on this fact
in https://github.com/mastodon/mastodon/pull/19903.)

```rb
a, b = 1, 2
Benchmark.ips do |bm|
  bm.report('conditional') { a > b ? a : b }
  bm.report('method') { [a, b].max }
  bm.compare!
end
```

Before:

```
Comparison:
         conditional: 22603733.2 i/s
              method: 19820412.7 i/s - 1.14x  (± 0.00) slower
```

This commit replaces the method lookup with a new CMP basic op, which
gives the examples above equivalent performance.

After:

```
Comparison:
              method: 24022466.5 i/s
         conditional: 23851094.2 i/s - same-ish: difference falls within
error
```

Relevant benchmarks show an improvement to Array#max and Array#min when
not using the optimized newarray_max instruction as well. They are
noticeably faster for small arrays with the relevant types, and the same
or maybe a touch faster on larger arrays.

```
$ make benchmark COMPARE_RUBY=<master@5958c305> ITEM=array_min
$ make benchmark COMPARE_RUBY=<master@5958c305> ITEM=array_max
```

The benchmarks added in this commit also look generally improved.

Co-authored-by: John Hawthorn <jhawthorn@github.com>

4 files changed, 57 insertions, 0 deletions

diff --git a/benchmark/array_sort_int.yml b/benchmark/array_sort_int.yml
new file mode 100644
index 0000000000..7b9027ebf7
--- /dev/null
+++ b/benchmark/array_sort_int.yml
@@ -0,0 +1,15 @@
+prelude: |
+  ary2 = 2.times.to_a.shuffle
+  ary10 = 10.times.to_a.shuffle
+  ary100 = 100.times.to_a.shuffle
+  ary1000 = 1000.times.to_a.shuffle
+  ary10000 = 10000.times.to_a.shuffle
+
+benchmark:
+  ary2.sort: ary2.sort
+  ary10.sort: ary10.sort
+  ary100.sort: ary100.sort
+  ary1000.sort: ary1000.sort
+  ary10000.sort: ary10000.sort
+
+loop_count: 10000
diff --git a/benchmark/enum_minmax.yml b/benchmark/enum_minmax.yml
new file mode 100644
index 0000000000..9d01731abb
--- /dev/null
+++ b/benchmark/enum_minmax.yml
@@ -0,0 +1,25 @@
+prelude: |
+  set2 = 2.times.to_a.shuffle.to_set
+  set10 = 10.times.to_a.shuffle.to_set
+  set100 = 100.times.to_a.shuffle.to_set
+  set1000 = 1000.times.to_a.shuffle.to_set
+  set10000 = 10000.times.to_a.shuffle.to_set
+
+benchmark:
+  set2.min: set2.min
+  set10.min: set10.min
+  set100.min: set100.min
+  set1000.min: set1000.min
+  set10000.min: set10000.min
+  set2.max: set2.max
+  set10.max: set10.max
+  set100.max: set100.max
+  set1000.max: set1000.max
+  set10000.max: set10000.max
+  set2.minmax: set2.minmax
+  set10.minmax: set10.minmax
+  set100.minmax: set100.minmax
+  set1000.minmax: set1000.minmax
+  set10000.minmax: set10000.minmax
+
+loop_count: 10000
diff --git a/benchmark/enum_sort.yml b/benchmark/enum_sort.yml
new file mode 100644
index 0000000000..6f26e748c6
--- /dev/null
+++ b/benchmark/enum_sort.yml
@@ -0,0 +1,15 @@
+prelude: |
+  set2 = 2.times.to_a.shuffle.to_set
+  set10 = 10.times.to_a.shuffle.to_set
+  set100 = 100.times.to_a.shuffle.to_set
+  set1000 = 1000.times.to_a.shuffle.to_set
+  set10000 = 10000.times.to_a.shuffle.to_set
+
+benchmark:
+  set2.sort_by: set2.sort_by { 0 }
+  set10.sort_by: set10.sort_by { 0 }
+  set100.sort_by: set100.sort_by { 0 }
+  set1000.sort_by: set1000.sort_by { 0 }
+  set10000.sort_by: set10000.sort_by { 0 }
+
+loop_count: 10000
diff --git a/benchmark/range_min.yml b/benchmark/range_min.yml
new file mode 100644
index 0000000000..9e60dd7308
--- /dev/null
+++ b/benchmark/range_min.yml
@@ -0,0 +1,2 @@
+benchmark:
+  - (1..10).min