Potentially address performance regression with Julia 1.9

[BioTurboNick]

There's a performance issue with Julia 1.9-rc. When using the constructors with a length determined at runtime, Julia falls back on dynamic dispatch.

using BenchmarkTools
using StaticArrays
const xx = [1; 2]
@noinline function e(x)
    k = length(x)
    SVector{k}(x)
end

@btime e(xx) 
# 1.9-rc1: 27.898 μs (31 allocations: 2.66 KiB)
# 1.8.5: 762.239 ns (10 allocations: 544 bytes)

According to JuliaLang/julia#49145, there may be changes in StaticArrays that would work around this issue.

[N5N3]

It should be noticed that this is caused by the @propagate_inbounds macro, which makes the constuctor inlined unstablely.
I tried some other constructor in convert.jl. Looks like construct_type is heavy enough to block the inliner so perhaps we only need this specific fix.

[BioTurboNick]

Sorry, what's the specific fix you have in mind? I'd do a PR but not quite sure what you mean.

[N5N3]

We have some similar pattern in convert.jl, but JuliaLang/julia#49145 (comment) is enough as all other constructors are not marked as @inline or @propagate_inbounds