From ec5742f4ab463d2316a6a3324f6e4552cfc94c15 Mon Sep 17 00:00:00 2001
From: shuwenn <47200617+alphabetc1@users.noreply.github.com>
Date: Wed, 8 Apr 2026 00:19:31 +0800
Subject: [PATCH] fix: Auto-correct page_size for Mamba no_buffer radix cache
 mode (#20538)

---
 python/sglang/srt/server_args.py | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/python/sglang/srt/server_args.py b/python/sglang/srt/server_args.py
index 74445c9cd..f6757a1bd 100644
--- a/python/sglang/srt/server_args.py
+++ b/python/sglang/srt/server_args.py
@@ -2204,6 +2204,13 @@ class ServerArgs:
                     == 0
                 ), f"For SSM models with extra buffer, either FLA_CHUNK_SIZE or page_size must be divisible by the other, got {FLA_CHUNK_SIZE=}, {self.page_size=}"
         elif not self.disable_radix_cache:  # no_buffer
+            if self.page_size is not None and self.page_size != 1:
+                logger.warning(
+                    f"{model_arch} with radix cache requires page_size=1 in the current "
+                    f"Mamba scheduling mode (no_buffer), but got {self.page_size}. "
+                    "Automatically setting page_size=1."
+                )
+                self.page_size = 1
             if self.speculative_algorithm is None:
                 logger.warning(
                     "Disabling overlap schedule since mamba no_buffer is not compatible with "