Refactor TranscodeUtf8 to allow trimming of Vector128<T>

[tannergooding]

This resolves#47860 by moving where Unsafe.SkipInit is called so we can trim out Vector128<short> in the default blazor_wasm template.

I logged dotnet/linker#1805 to track the linker not understanding things like SkipInit (which are NonVersionable) and dotnet/linker#1808 to track the linker not trimming out a local when its only usage is it being initialized via initobj

CC. @eerhardt, @stephentoub

[ghost]

I couldn't figure out the best area label to add to this PR. If you have write-permissions please help me learn by adding exactly one area label.

[GrabYourPitchforks]

TBH we should probably have a separate implementation of the UTF-8 logic for interpreted or wasm scenarios. The current validation and transcoding logic is heavily optimized to take into consideration things like branch prediction, cache locality, and efficient register utilization. This results in the methods being large and complex because they consist of a bunch of little worker loops that each work on one optimized scenario.

A separate interpreted / wasm-specific implementation would be a straightforward "just do the simplest thing" loop. No cleverness, just a small simple loop that consists of easy instructions.

Edit - an example of the "simple" validation and transcoding loops:

// simplest validation loopintGetIndexOfFirstInvalidUtf8Byte(ROS<byte>input,outintutf16CharCount){intoriginalInputLength=input.Length;utf16CharCount=0;while(!input.IsEmpty){if(Rune.DecodeUtf8(input,outRunerune,outintbytesConsumed)!=OperationStatus.Done)break;input=input.Slice(bytesConsumed);utf16CharCount+=input.Utf16CodeUnitCount;}returnoriginalInputLength-input.Length;}// simplest transcoding loopOperationStatusTranscodeUtf8ToUtf16(ROS<byte>input,ROS<char>output,outintbytesConsumed,outintcharsWritten){intoriginalInputLength=input.Length;intoriginalOutputLength=output.Length;OperationStatusopStatus=OperationStatus.Done;while(!input.IsEmpty){opStatus=Rune.DecodeUtf8(input,outRunerune,outintbytesConsumedJustNow);if(opStatus!=OperationStatus.Done){break;}if(!rune.TryWriteUtf16(output,outintcharsWrittenJustNow)){opStatus=OperationStatus.DestinationTooSmall;break;}input=input.Slice(bytesConsumedJustNow);output=output.Slice(charsWrittenJustNow);}bytesConsumed=originalInputLength-input.Length;charsWritten=originalOutputLength-output.Length;returnopStatus;}

[ghost]

Tagging subscribers to 'linkable-framework': @eerhardt, @vitek-karas, @LakshanF, @tannergooding, @sbomer
See info in area-owners.md if you want to be subscribed.

Issue Details

---

This resolves #47860 by moving where Unsafe.SkipInit is called so we can trim out Vector128<short> in the default blazor_wasm template.

I logged dotnet/linker#1805 to track the linker not understanding things like SkipInit (which are NonVersionable) and dotnet/linker#1808 to track the linker not trimming out a local when its only usage is it being initialized via initobj

CC. @eerhardt, @stephentoub

Author:	tannergooding
Assignees:	-
Labels:	area-System.Text.Encoding, linkable-framework
Milestone:	-

[eerhardt]

Nice! LGTM

[tannergooding]

@GrabYourPitchforks, could you log an "up-for-grabs" issue suggesting the above?

[ghost]

Hello @tannergooding!

Because this pull request has the auto-merge label, I will be glad to assist with helping to merge this pull request once all check-in policies pass.

p.s. you can customize the way I help with merging this pull request, such as holding this pull request until a specific person approves. Simply @mention me (@msftbot) and give me an instruction to get started! Learn more here.