Refactor how SwitchInt stores jump targets

[jonas-schievink]

Closes#65693

[rust-highfive]

r? @eddyb

(rust_highfive has picked a reviewer for you, use r? to override)

[jonas-schievink]

@bors try @rust-timer queue

r? @oli-obk

[rust-timer]

Awaiting bors try build completion

[bors]

⌛ Trying commit 432535d with merge f75713278ef412b926ff605fed784633ba5d6092...

[bors]

☀️ Try build successful - checks-actions, checks-azure
Build commit: f75713278ef412b926ff605fed784633ba5d6092 (f75713278ef412b926ff605fed784633ba5d6092)

[rust-timer]

Queued f75713278ef412b926ff605fed784633ba5d6092 with parent cae8bc1, future comparison URL.

[rust-timer]

Finished benchmarking try commit (f75713278ef412b926ff605fed784633ba5d6092): comparison url.

Benchmarking this pull request likely means that it is perf-sensitive, so we're automatically marking it as not fit for rolling up. Please note that if the perf results are neutral, you should likely undo the rollup=never given below by specifying rollup- to bors.

Importantly, though, if the results of this run are non-neutral do not roll this PR up -- it will mask other regressions or improvements in the roll up.

@bors rollup=never

[jonas-schievink]

Looks like no noticeable change so far, good. Let's try using SmallVec instead of the old setup.

[jonas-schievink]

@bors try @rust-timer queue

[rust-timer]

Awaiting bors try build completion

[bors]

⌛ Trying commit 9a47f74 with merge 70f6dc489a31b5239cdff8395134265d0d6452cc...

[bors]

☀️ Try build successful - checks-actions, checks-azure
Build commit: 70f6dc489a31b5239cdff8395134265d0d6452cc (70f6dc489a31b5239cdff8395134265d0d6452cc)

[rust-timer]

Queued 70f6dc489a31b5239cdff8395134265d0d6452cc with parent 790d19c, future comparison URL.

[rust-timer]

Finished benchmarking try commit (70f6dc489a31b5239cdff8395134265d0d6452cc): comparison url.

Benchmarking this pull request likely means that it is perf-sensitive, so we're automatically marking it as not fit for rolling up. Please note that if the perf results are neutral, you should likely undo the rollup=never given below by specifying rollup- to bors.

Importantly, though, if the results of this run are non-neutral do not roll this PR up -- it will mask other regressions or improvements in the roll up.

@bors rollup=never

[jonas-schievink]

Hmm, instructions went down, but cycles and clocks went up. Not sure if that's just noise.

[oli-obk]

cycles and clocks are always the least reliable. The change lgtm and without you mentioning it I wouldn't even have looked at anything but instructions.

@bors r+

[bors]

📌 Commit 9a47f74 has been approved by oli-obk

[bors]

⌛ Testing commit 9a47f74 with merge afb4514...

[bors]

☀️ Test successful - checks-actions, checks-azure
Approved by: oli-obk
Pushing afb4514 to master...

[Mark-Simulacrum]

Perf results after landing are about as expected; I agree with @oli-obk that I would not want to change anything here and I think investigating is not warranted to prioritize right now.