From 6fe0f09c175e77a0e5e2c2882f5f73afb8a22fcf Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Johannes=20G=C3=A4=C3=9Fler?= <johannesg@5d6.de>
Date: Fri, 9 May 2025 23:31:46 +0200
Subject: [PATCH] CUDA: fix FlashAttention on Turing

---
 ggml/src/ggml-cuda/fattn-mma-f16.cuh | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/ggml/src/ggml-cuda/fattn-mma-f16.cuh b/ggml/src/ggml-cuda/fattn-mma-f16.cuh
index 2b6bdc30c024b..b2f95fa3f00e6 100644
--- a/ggml/src/ggml-cuda/fattn-mma-f16.cuh
+++ b/ggml/src/ggml-cuda/fattn-mma-f16.cuh
@@ -546,7 +546,7 @@ static __device__ __forceinline__ void flash_attn_ext_f16_iter(
         const int i0_stop = i0_start + 2*c::nbatch_V2 < DV ? i0_start + 2*c::nbatch_V2 : DV;
         const int i0_diff = i0_stop - i0_start;
 
-        if (nstages == 1) {
+        if (nstages <= 1) {
             constexpr bool use_cp_async = nstages == 1;
             flash_attn_ext_f16_load_tile<stride_tile_V, nwarps, c::nbatch_fa, use_cp_async>
                 (V_h2 + k_VKQ_0*stride_V + i0_start/2, tile_V, i0_diff/2, stride_V);