adjust Miri to needs of changed unwinding strategy

[RalfJung]

As expected, #67502 broke unwinding in Miri. To fix it we have to adjust parts of the engine and the panic runtime, which this PR does. The Miri-side changes are in rust-lang/miri#1227.

Cc @oli-obk @Aaron1011 @Mark-Simulacrum @Amanieu

[rust-highfive]

r? @davidtwco

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

[RalfJung]

@bors p=1
because we cannot even test Miri with rustc master until this lands (so there is a risk of other breakage landing, and the more that accumulates the harder it is to fix)

[davidtwco]

r? @oli-obk

[Mark-Simulacrum]

Seems good, though I wasn't able to quite follow the miri changes.

[Mark-Simulacrum]

These seem like they could be separate? But fine to leave them in too.

[RalfJung]

I had to add from_i32 for the Miri side of this. So while I was at it I figured I might as well add the comments here, instead of making them a separate PR.

[oli-obk]

r=me with the comment updated.

[RalfJung]

@bors r=oli-obk

[bors]

📌 Commit b450e1b has been approved by oli-obk

[bors]

⌛ Testing commit b450e1b with merge 7cdbc87...

[bors]

☀️ Test successful - checks-azure
Approved by: oli-obk
Pushing 7cdbc87 to master...