Skip to content

LLM inference in C/C++

C++ ggml

cuda: fix warnings for unused variable/function (#26688)

David Friehs5b87ed30f84a
diff --git a/ggml/src/ggml-cuda/ggml-cuda.cu b/ggml/src/ggml-cuda/ggml-cuda.cu
index f4b271146c..5446b31318 100644
--- a/ggml/src/ggml-cuda/ggml-cuda.cu
+++ b/ggml/src/ggml-cuda/ggml-cuda.cu
@@ -5209,6 +5209,7 @@ static bool ggml_backend_cuda_device_offload_op(ggml_backend_dev_t dev, const gg
static ggml_backend_event_t ggml_backend_cuda_device_event_new(ggml_backend_dev_t dev) {
#ifdef GGML_CUDA_NO_PEER_COPY
+ GGML_UNUSED(dev);
return nullptr;
#else
ggml_backend_cuda_device_context * dev_ctx = (ggml_backend_cuda_device_context *)dev->context;
diff --git a/ggml/src/ggml-cuda/quantize.cu b/ggml/src/ggml-cuda/quantize.cu
index 2bd9b62623..bcc7723957 100644
--- a/ggml/src/ggml-cuda/quantize.cu
+++ b/ggml/src/ggml-cuda/quantize.cu
@@ -8,7 +8,6 @@ struct __builtin_align__(32) float8 {
float x; float y; float z; float w;
float p; float q; float r; float s;
};
-#endif
#if CUDART_VERSION >= 12080
static __device__ __forceinline__ float nvfp4_native_scale_error(
@@ -49,6 +48,7 @@ static __device__ __forceinline__ float nvfp4_native_scale_error(
return err;
}
#endif // CUDART_VERSION >= 12080
+#endif // defined(BLACKWELL_MMA_AVAILABLE)
__launch_bounds__(CUDA_QUANTIZE_BLOCK_SIZE, 1)
static __global__ void quantize_q8_1(