From e98f9ac554bd3e983da68ac1fa64c2c31c04686e Mon Sep 17 00:00:00 2001
From: AidanBeltonS <87009434+AidanBeltonS@users.noreply.github.com>
Date: Mon, 27 May 2024 17:34:51 +0100
Subject: [PATCH] Fix q_xxs using mul_mat_q (llama/7459)

---
 ggml-sycl.cpp | 1 +
 1 file changed, 1 insertion(+)

diff --git a/ggml-sycl.cpp b/ggml-sycl.cpp
index f329bc2..8839f77 100644
--- a/ggml-sycl.cpp
+++ b/ggml-sycl.cpp
@@ -15263,6 +15263,7 @@ static void ggml_sycl_mul_mat(const ggml_tensor * src0, const ggml_tensor * src1
             }
         } else {
             bool use_mul_mat_q = min_compute_capability >= VER_4VEC && ggml_is_quantized(src0->type);
+            use_mul_mat_q = use_mul_mat_q && (src0->type != GGML_TYPE_IQ2_XXS);
 
             if (use_xmx && min_compute_capability >= VER_GEN9 && src1->ne[1] > XMX_MAX_BATCH_SIZE) {
                 use_mul_mat_q = false;