Release v2.6.2

[lkdvos]

A performance-focused patch release that reduces compilation time and with small improvements to the mapreduce kernels.

Highlights

• Faster compilation: @nospecialize added to the mapreduce dispatch chain (Add @nospecialize to mapreduce dispatch chain to cut compile time #66) and blas_mul! de-specialized (Despecialize blas_mul! for improved compilation time #73) to cut compile time.
• Faster reductions: added a unit-stride fast path to _mapreduce_kernel! and specialized the non-reducing kernel by dropping the reduction-only iszero hoist (Specialize non-reducing kernel: drop reduction-only iszero hoist #69).
• Kernel ordering: improved loop ordering in the mapreduce kernel — "order → fuse → order" (order - fuse - order kernel logic #74).
• Minor internal cleanups (simplified firststride condition, improved comments) and CI maintenance (bump actions/checkout 6 → 7, Bump actions/checkout from 6 to 7 #68).

Full Changelog

See the commit history for the complete list of changes.

🤖 Generated with Claude Code

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.