Simplify encoder and decoder

[cjgillot]

Extracted from #83036 and #82780.

[rust-highfive]

r? @lcnr

(rust-highfive has picked a reviewer for you, use r? to override)

[joshtriplett]

@bors try @rust-timer queue

[rust-timer]

Awaiting bors try build completion.

@rustbot label: +S-waiting-on-perf

[bors]

⌛ Trying commit ccef537b0366d1ced16cf212d385aa1707a5794f with merge bf696c7a499ba76acd374ad810be4e05f17460b9...

[bors]

☀️ Try build successful - checks-actions
Build commit: bf696c7a499ba76acd374ad810be4e05f17460b9 (bf696c7a499ba76acd374ad810be4e05f17460b9)

[rust-timer]

Queued bf696c7a499ba76acd374ad810be4e05f17460b9 with parent 9d0446f, future comparison URL.

[rust-timer]

Finished benchmarking try commit (bf696c7a499ba76acd374ad810be4e05f17460b9): 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
@rustbot label: +S-waiting-on-review -S-waiting-on-perf

[joshtriplett]

This provides a decent improvement on max-rss, but regresses on instructions.

[cjgillot]

@bors try @rust-timer queue

[rust-timer]

Awaiting bors try build completion.

@rustbot label: +S-waiting-on-perf

[bors]

⌛ Trying commit 637253f5263e9d8abf2b522553c64ab80d53f358 with merge 0f5649699bc49e9345864d7b99c11697dd8506f3...

[bors]

☀️ Try build successful - checks-actions
Build commit: 0f5649699bc49e9345864d7b99c11697dd8506f3 (0f5649699bc49e9345864d7b99c11697dd8506f3)

[rust-timer]

Queued 0f5649699bc49e9345864d7b99c11697dd8506f3 with parent 1705a7d, future comparison URL.

[rust-timer]

Finished benchmarking try commit (0f5649699bc49e9345864d7b99c11697dd8506f3): 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
@rustbot label: +S-waiting-on-review -S-waiting-on-perf

[joshtriplett]

Still appears to be a substantial regression, unfortunately.

[michaelwoerister]

I've done a few similar things in c49297d (part of #82183). I did not see performance regressions there. Maybe that's of interest.

[rust-timer]

Awaiting bors try build completion.

@rustbot label: +S-waiting-on-perf

[bors]

⌛ Trying commit 11b3409 with merge 58740d89a14c455cd96c58a6dc4a3478d8cb854c...

[bors]

☀️ Try build successful - checks-actions
Build commit: 58740d89a14c455cd96c58a6dc4a3478d8cb854c (58740d89a14c455cd96c58a6dc4a3478d8cb854c)

[rust-timer]

Queued 58740d89a14c455cd96c58a6dc4a3478d8cb854c with parent 9f4bc3e, future comparison URL.

[rust-timer]

Finished benchmarking try commit (58740d89a14c455cd96c58a6dc4a3478d8cb854c): 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
@rustbot label: +S-waiting-on-review -S-waiting-on-perf

[bors]

@rust-timer: 🔑 Insufficient privileges: not in try users

[Aaron1011]

cc @rust-lang/infra ^ - it looks like the permissions for rust-timer are messed up somehow.

[bjorn3]

Improvements up to 1.5%.

[lcnr]

can someone else take this over, I don't have a lot of capacity rn

r? @michaelwoerister maybe, feel free to reassign

[michaelwoerister]

Looks good to me. I'm still not a fan of all that unsafe code and most of it can be removed as c49297d demonstrates.

I'd also prefer if read_raw_bytes would just borrow the underlying data instead of copying it to &mut [MaybeUninit<u8>]. That works with the opaque decoder but can't be done if read_raw_bytes is moved to the Decoder trait, as this PR does. I hope we can make that change once we get rid of JSON encoding as rust-lang/compiler-team#418 proposes.

But for now this PR is still a strict improvement.

@bors r+

[bors]

📌 Commit 11b3409 has been approved by michaelwoerister

[bors]

⌛ Testing commit 11b3409 with merge d04c3aa...

[bors]

☀️ Test successful - checks-actions
Approved by: michaelwoerister
Pushing d04c3aa to master...