From cb6caca191b9a3a9a4eaa13dd9e465225d127034 Mon Sep 17 00:00:00 2001 From: Neo Zhang Date: Fri, 23 Jan 2026 20:54:10 +0800 Subject: [PATCH] [SYCL] use malloc to support both iGPU and dGPU in same time (#18992) * use malloc to support both iGPU and dGPU in same time * support windows --------- Co-authored-by: Neo Zhang Jianyu --- ggml/src/ggml-sycl/ggml-sycl.cpp | 21 ++++++++++++++++++--- 1 file changed, 18 insertions(+), 3 deletions(-) diff --git a/ggml/src/ggml-sycl/ggml-sycl.cpp b/ggml/src/ggml-sycl/ggml-sycl.cpp index bb8acc922b..ce2f0d41c9 100644 --- a/ggml/src/ggml-sycl/ggml-sycl.cpp +++ b/ggml/src/ggml-sycl/ggml-sycl.cpp @@ -1157,13 +1157,28 @@ static const char * ggml_backend_sycl_host_buffer_type_name(ggml_backend_buffer_ GGML_UNUSED(buft); } +inline void * aligned_malloc_host(size_t alignment, size_t size) { +#ifdef _WIN32 + return _aligned_malloc(size, alignment); +#else + return aligned_alloc(alignment, size); +#endif +} + +inline void free_aligned_mem_host(void * memblock) { +#ifdef _WIN32 + _aligned_free(memblock); +#else + free(memblock); +#endif +} + static void ggml_backend_sycl_host_buffer_free_buffer(ggml_backend_buffer_t buffer) { - ggml_sycl_host_free(buffer->context); + free_aligned_mem_host((void *)buffer->context); } static ggml_backend_buffer_t ggml_backend_sycl_host_buffer_type_alloc_buffer(ggml_backend_buffer_type_t buft, size_t size) { - void * ptr = ggml_sycl_host_malloc(size); - + void * ptr = aligned_malloc_host(TENSOR_ALIGNMENT, size); if (ptr == nullptr) { // fallback to cpu buffer return ggml_backend_buft_alloc_buffer(ggml_backend_cpu_buffer_type(), size);