Regeneralize the "-1 for None" change

[scottmcm]

Not planning on un-drafting this until after #158734 lands.

The "prefer the smaller magnitude" change (#155473) had been applied pretty narrowly: just to option-like niches and just to non-bool-like values. But now that llvm/llvm-project#208611 has been fixed, let's apply it more generally.

In doing so I ended up pretty substantially restructuring the function. I also greatly expanded the comments and added unit tests for it so it's easier to exercise all the paths (and much quicker than rebuilding all of rustc). I made sure to keep the previous "try to keep zero available for None" cleverness, however.

---

Or, if you prefer, this is contradicting the old comment in the code that

If niche zero is already reserved, the selection of bounds are of little interest.

🙂

[scottmcm]

:/

sccache: error: Timed out waiting for server startup. Maybe the remote service is unreachable?

[scottmcm]

@bors try jobs=aarch64llvm-21

[rust-bors]

💔 Test for 84c2f57 failed: CI. Failed jobs:

• try - aarch64-gnu-llvm-21-1 (web logs, enhanced plaintext logs)
• try - aarch64-gnu-llvm-21-2 (web logs, enhanced plaintext logs)

[scottmcm]

@bors try @rust-timer queue

[rust-bors]

☀️ Try build successful (CI)
Build commit: 99c49b3 (99c49b36c0e329e5984e2b569f71439d947c60e2)
Base parent: 87212ce (87212cef77e7bfa92ba0c4850be1d089745ba6fc)

[rust-timer]

Finished benchmarking commit (99c49b3): comparison URL.

Overall result: ❌✅ regressions and improvements - please read:

Benchmarking means the PR may be perf-sensitive. It's automatically marked not fit for rolling up. Overriding is possible but disadvised: it risks changing compiler perf.

Next, please: If you can, justify the regressions found in this try perf run in writing along with @rustbot label: +perf-regression-triaged. If not, fix the regressions and do another perf run. Neutral or positive results will clear the label automatically.

@bors rollup=never rustc-perf
@rustbot label: -S-waiting-on-perf +perf-regression

Instruction count

Our most reliable metric. Used to determine the overall result above. However, even this metric can be noisy.

	mean	range	count
Regressions ❌ (primary)	0.3%	[0.3%, 0.3%]	1
Regressions ❌ (secondary)	1.3%	[0.3%, 2.4%]	25
Improvements ✅ (primary)	-0.2%	[-0.3%, -0.2%]	12
Improvements ✅ (secondary)	-0.7%	[-2.8%, -0.2%]	20
All ❌✅ (primary)	-0.2%	[-0.3%, 0.3%]	13

Max RSS (memory usage)

Results (primary 3.9%, secondary 26.9%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	3.9%	[3.9%, 3.9%]	1
Regressions ❌ (secondary)	26.9%	[23.4%, 30.5%]	12
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	3.9%	[3.9%, 3.9%]	1

Cycles

Results (primary -3.0%, secondary 16.7%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	19.8%	[7.4%, 31.8%]	12
Improvements ✅ (primary)	-3.0%	[-3.0%, -3.0%]	1
Improvements ✅ (secondary)	-2.3%	[-2.5%, -2.0%]	2
All ❌✅ (primary)	-3.0%	[-3.0%, -3.0%]	1

Binary size

Results (primary -0.1%, secondary -0.0%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	0.0%	[0.0%, 0.0%]	3
Regressions ❌ (secondary)	0.0%	[0.0%, 0.0%]	11
Improvements ✅ (primary)	-0.1%	[-0.1%, -0.1%]	13
Improvements ✅ (secondary)	-0.8%	[-0.8%, -0.8%]	1
All ❌✅ (primary)	-0.1%	[-0.1%, 0.0%]	16

Bootstrap: 456.607s -> 458.361s (0.38%)
Artifact size: 399.19 MiB -> 399.08 MiB (-0.03%)