docs: clarify v2.9.0 is unofficial fork + add GitHub plan/notes links

README.md

@@ -27,6 +27,13 @@ Pre-built version: lash_attn-2 with **Python 3.11 & 3.12 & 3.13** support.
 - 15.11.2025 Uploaded v2.8.3 based on PyTorch 2.9.1+cu130
 - 12.02.2026 Uploaded v2.8.3 based on PyTorch 2.10.0+cu130
 - 29.03.2026 Uploaded v2.8.3 based on PyTorch 2.11.0+cu130
-- 13.05.2026 Uploaded v2.9.0 based on PyTorch 2.10.0+cu130 (FA2 fork with A-1/A-2 optimizations)
+- 13.05.2026 Uploaded **v2.9.0** based on PyTorch 2.10.0+cu130 — **unofficial fork-only build** (not an official release). Includes FA2 A-1/A-2 optimizations.
+
+## About v2.9.0
+**v2.9.0 is not an official FlashAttention release.** It is an independent fork build that continues FA2 kernel development while upstream focuses on FA3/F4.
+
+- **Optimization plan (GitHub):** https://github.com/ussoewwin/flash-attention/blob/main/AI/FA2_BACKPORT_FROM_FA4_PLAN.md
+- **Kernel change notes (GitHub):** https://github.com/ussoewwin/flash-attention/blob/main/md/FA2_CHANGES_v1.1.md
+- **Fork changelog (GitHub):** https://github.com/ussoewwin/flash-attention/blob/main/md/CHANGELOG.md
 
 ※Unofficial built version!! It works correctly in my environment, but I am not sure that will work in yours.