From: Georgi Gerganov <redacted>
Date: Wed, 22 May 2024 09:36:37 +0000 (+0300)
Subject: cuda : fix compile warning (llama/7454)
X-Git-Tag: upstream/0.0.1642~656
X-Git-Url: https://git.djapps.eu/?a=commitdiff_plain;h=cedea395fe62f400c2813fd25ef03f50d49ded80;p=pkg%2Fggml%2Fsources%2Fggml

cuda : fix compile warning (llama/7454)
---

diff --git a/src/ggml-cuda/fattn-tile-f32.cu b/src/ggml-cuda/fattn-tile-f32.cu
index 54db765e..b8b2f69e 100644
--- a/src/ggml-cuda/fattn-tile-f32.cu
+++ b/src/ggml-cuda/fattn-tile-f32.cu
@@ -283,8 +283,7 @@ void launch_fattn_tile_f32_64_128(ggml_backend_cuda_context & ctx, ggml_tensor *
 }
 
 void ggml_cuda_flash_attn_ext_tile_f32(ggml_backend_cuda_context & ctx, ggml_tensor * dst) {
-    const ggml_tensor * KQV = dst;
-    const ggml_tensor * Q   = dst->src[0];
+    const ggml_tensor * Q = dst->src[0];
 
     if (Q->ne[1] <= 16) {
         constexpr int cols_per_block = 16;