Implement Vector128 version of System.Buffers.Text.Base64 DecodeFromUtf8 and EncodeTo...

Implement Vector128 version of System.Buffers.Text.Base64 DecodeFromUtf8 and EncodeToUtf8 (#70654)

* Implement Vector128 version of System.Buffers.Text.Base64.DecodeFromUtf8

Rework the SS3 into a Vector128 version, and add Arm64 support.

* SSE3 improvements

* Remove superfluous bitwise And

* Add comment to SimdShuffle

* Inline SimdShuffle

* Implement Vector128 version of System.Buffers.Text.Base64.EncodeToUtf8

* Ensure masking on SSE3

Change-Id: I319f94cfc51d0542ae4eb11a8d48b3eb8180553f
CustomizedGitHooks: yes

* Restore asserts and move zero inside the loop

* Neater C# code

Change-Id: I2cbe14f4228f8035e7d213b5b58815c4eee35563
CustomizedGitHooks: yes

* Make SimdShuffle consistent across X64 and Arm64

* Better looking multiply