From 156d1af23a4fa80809ab423bc1042d732ef6f95f Mon Sep 17 00:00:00 2001
From: Chetan Kumar Verma <39086835+ckvermaAI@users.noreply.github.com>
Date: Tue, 26 May 2026 08:32:21 +0530
Subject: [PATCH] [Intel GPU] Fix incorrect KV-cache page table for local
 attention when page_size > 1 (#23757)

---
 python/sglang/srt/layers/attention/xpu_backend.py | 11 +++++++++++
 1 file changed, 11 insertions(+)

diff --git a/python/sglang/srt/layers/attention/xpu_backend.py b/python/sglang/srt/layers/attention/xpu_backend.py
index c7149c1a1..3b5743799 100644
--- a/python/sglang/srt/layers/attention/xpu_backend.py
+++ b/python/sglang/srt/layers/attention/xpu_backend.py
@@ -925,6 +925,17 @@ class XPUAttentionBackend(AttentionBackend):
             metadata.local_attn_metadata = None
             return
 
+        # make_local_attention_virtual_batches expects a page-granularity block table:
+        # column p is the logical page number, and the value stored at that column is the
+        # physical page index. The raw req_to_token table is token-granularity (column i =
+        # the KV slot for token i), so when page_size > 1 we must stride and divide first
+        # so that block_starts = k_seqstarts_absolute // page_size correctly indexes the table.
+        if self.page_size > 1:
+            strided_indices = torch.arange(
+                0, page_table.shape[1], self.page_size, device=page_table.device
+            )
+            page_table = page_table[:, strided_indices] // self.page_size
+
         cu_seqlens_q_np = cu_seqlens_q.cpu().numpy()
         seq_lens_np = cache_seqlens_int32.cpu().numpy()
         (