Fix q_xxs using mul_mat_q (llama/7459)

diff --git a/src/ggml-sycl.cpp b/src/ggml-sycl.cpp
index f329bc27265feab2cd2a1b0b806bb35e050286a9..8839f775d5b880c507ef5d8efa96ac20497d2eb7 100644 (file)
--- a/src/ggml-sycl.cpp
+++ b/src/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;