all: replace math/rand with math/rand/v2

[seankhliao]

Update to new stdlib apis.

The new Float64 is uniform, which resolves a long comment.

https://cs.opensource.google/go/go/+/refs/tags/go1.24.2:src/math/rand/v2/rand.go;l=209

// There are exactly 1<<53 float64s in [0,1). Use Intn(1<<53) / (1<<53).
return float64(r.Uint64()<<11>>11) / (1 << 53)

goos: linux
goarch: amd64
pkg: go.opentelemetry.io/otel/sdk/trace
cpu: 12th Gen Intel(R) Core(TM) i7-1260P
                                 │   old.txt    │               new.txt                │
                                 │    sec/op    │    sec/op     vs base                │
TraceStart/with_a_simple_span-16   387.1n ±  5%   306.8n ± 15%  -20.73% (p=0.000 n=10)
TraceStart/with_several_links-16   542.2n ±  5%   501.0n ±  2%   -7.61% (p=0.000 n=10)
TraceStart/with_attributes-16      521.4n ± 14%   571.6n ±  6%   +9.64% (p=0.009 n=10)
geomean                            478.3n         444.6n         -7.05%

                                 │  old.txt   │               new.txt               │
                                 │    B/op    │    B/op     vs base                 │
TraceStart/with_a_simple_span-16   528.0 ± 0%   528.0 ± 0%       ~ (p=1.000 n=10) ¹
TraceStart/with_several_links-16   704.0 ± 0%   704.0 ± 0%       ~ (p=1.000 n=10) ¹
TraceStart/with_attributes-16      784.0 ± 0%   784.0 ± 0%       ~ (p=1.000 n=10) ¹
geomean                            663.0        663.0       +0.00%
¹ all samples are equal

                                 │  old.txt   │               new.txt               │
                                 │ allocs/op  │ allocs/op   vs base                 │
TraceStart/with_a_simple_span-16   2.000 ± 0%   2.000 ± 0%       ~ (p=1.000 n=10) ¹
TraceStart/with_several_links-16   3.000 ± 0%   3.000 ± 0%       ~ (p=1.000 n=10) ¹
TraceStart/with_attributes-16      4.000 ± 0%   4.000 ± 0%       ~ (p=1.000 n=10) ¹
geomean                            2.884        2.884       +0.00%
¹ all samples are equal

[pellared]

Are we sure that we do not want a changelog entry for this change?

[seankhliao]

I thought about it but it didn't seem to fit the rest of the changelog as there's no real observable change from the end user perspective.

[pellared]

I thought about it but it didn't seem to fit the rest of the changelog as there's no real observable change from the end user perspective.

What about the fact that FixedSizeReservoir now uses a uniform random generator?
Is it not a fix/change so that we are going to be compliant with https://github.com/open-telemetry/opentelemetry-specification/blob/cdac03154c5b814a83689aecb97f82d6bbf7c0b2/specification/metrics/sdk.md?plain=1#L1190-L1192?

[MrAlias]

Looks good overall. I especially like the ID generator fixes.

Please include benchmarks using benchstat in the PR descrition for the files changed that provide existing benchmarks.

[MrAlias]

🚀

[seankhliao]

I added a benchstat comparison for the TraceStart benchmarks, not sure if anything else would be affected?

[dmathieu]

Unless anyone objects, I will merge this in about 24 hours.