LATX: optimize AVX and scalar instruction sequences

[luzeng87]

Summary

• combine related AVX shuffle, insert, compare, and arithmetic sequence optimizations
• recognize repeated scalar additions and three-vector sums only when register use proves the rewrite safe
• forward scalar arithmetic into FMA operations with explicit data-flow checks
• preserve VCOMIS/VUCOMIS ordered and unordered flag behavior in translated and LBT paths
• fuse a guarded scalar pixel update sequence
• translate VROUNDPS SAE truncation without serialized FCSR reads and writes
• preserve VINSERTPS inputs when destination and both sources alias

Organization

This PR is stacked on #437. Its branch contains the five #437 prerequisite commits followed by one squashed instruction-sequence commit. Review the final commit for the additions in this PR. The previous cumulative audio and pixel-update PRs are superseded by this grouped PR.

Validation

• instruction tests cover source/destination aliases, intermediate-register reads, NaN, infinity, signed zero, subnormal values, and VCOMIS unordered cases
• fixtures run in JIT, cold AOT, and hot AOT modes
• test and file names describe instructions rather than external workloads
• DCO sign-offs present on all commits

This PR remains a draft while the grouped branch checks run.

[xiezyang]

VINSERTPS 快路径未处理 dest == src1 == src2 的别名情况。
例如 vinsertps xmm4, xmm4, xmm4, 0x9c 会在读取 src1 前覆盖 dest，导致原始 lane 0 丢失。
建议该快路径排除 dest_is_src1，或先保存 src1 后再写目标寄存器。

[xiezyang]

本 PR 包含的 b245da0 存在 3 处 YMM 结果未写回问题：

• VMASKMOVPS/PD
• VPBLENDVB
• VPBLENDW（立即数不是 0 或 0xff 时）
  这些路径均将结果计算到 temp，但只在 XMM 形式下写回 dest。因此 YMM 目标寄存器低 128 位保留旧值，高 128 位被错误清零。
  建议无条件将 temp 写回 dest，并且仅在 XMM 目标时调用 set_high128_xreg_to_zero(dest)。

[luzeng87]

This stack inherits the reproducible mixed XMM/YMM failure from #437 and also has the VINSERTPS destination/source alias issue identified in review. The independent changes need to be submitted and verified by function, without this shared stack. Closing this PR.

[luzeng87]

Synced the YMM mask/blend writeback correction as commit dccd7ee83d. This branch had embedded the earlier deferred-zeroing commit, so it required the same full-width destination fix and regression test. The cumulative branch containing this patch passed the focused JIT/cold-AOT/hot-AOT test and lat-pr-fast 24/24 on LoongArch 3A6000. This PR remains a draft.

[luzeng87]

Reorganized this branch on top of the current #437 head. All sequence changes are now represented by one squashed commit, e4403b032b; the former cumulative audio and pixel-update PRs are superseded. The VINSERTPS dest == src1 == src2 review case is covered and passes in JIT, cold AOT, and hot AOT modes. Test names now describe instructions and transformations only.