From 9d0693bce38013364b1042568d9083353bfff48f Mon Sep 17 00:00:00 2001 From: kiltyj Date: Mon, 5 Jun 2023 13:24:04 -0700 Subject: metal : use shared buffers between CPU and GPU (#1696) * Use MTLDevice.newBufferWithBytesNoCopy to share buffers between CPU and GPU * Page-align buffers used by Metal * Remove trailing whitespace * Only import unistd.h for Metal builds * metal : remove unnecessary copies --------- Co-authored-by: Georgi Gerganov --- llama-util.h | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) (limited to 'llama-util.h') diff --git a/llama-util.h b/llama-util.h index 3cac9f68..4f8a4296 100644 --- a/llama-util.h +++ b/llama-util.h @@ -405,13 +405,29 @@ struct llama_buffer { llama_buffer() = default; void resize(size_t len) { +#ifdef GGML_USE_METAL + free(addr); + int result = posix_memalign((void **) &addr, getpagesize(), len); + if (result == 0) { + memset(addr, 0, len); + } + else { + addr = NULL; + } +#else delete[] addr; addr = new uint8_t[len]; +#endif size = len; } ~llama_buffer() { +#ifdef GGML_USE_METAL + free(addr); +#else delete[] addr; +#endif + addr = NULL; } // disable copy and move -- cgit v1.2.3