add back LLVM-specific funnel shift implementation

[RalfJung]

This reverts commit a00db66. That commit is likely the culprit for the unexpected perf regression in #148478; let's see if reverting it helps.

[RalfJung]

@bors try
@rust-timer queue

[rust-bors]

☀️ Try build successful (CI)
Build commit: be9fa2b (be9fa2b38b9d60137dc04dafbb163043ca4ccaf7, parent: cf8a95590a1b768b7711f2631d5b5e3ead464de7)

[rust-timer]

Finished benchmarking commit (be9fa2b): comparison URL.

Overall result: ✅ improvements - no action needed

Benchmarking this pull request means it may be perf-sensitive – we'll automatically label it not fit for rolling up. You can override this, but we strongly advise not to, due to possible changes in compiler perf.

@bors rollup=never
@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
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.2%	[-0.3%, -0.2%]	7
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

This benchmark run did not return any relevant results for this metric.

Cycles

Results (primary -2.4%, secondary 0.4%)

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)	2.7%	[2.6%, 2.7%]	2
Improvements ✅ (primary)	-2.4%	[-2.7%, -2.3%]	3
Improvements ✅ (secondary)	-4.1%	[-4.1%, -4.1%]	1
All ❌✅ (primary)	-2.4%	[-2.7%, -2.3%]	3

Binary size

This benchmark run did not return any relevant results for this metric.

Bootstrap: 470.335s -> 472.107s (0.38%)
Artifact size: 386.93 MiB -> 386.95 MiB (0.00%)

[RalfJung]

So this does actually help a bit but only for a synthetic benchmark... doesn't seem worth it.