From a50d0033020daa3c5ef921f0bde9569a9929c9af Mon Sep 17 00:00:00 2001 From: JohannesGaessler Date: Thu, 8 Jun 2023 11:21:51 +0200 Subject: [PATCH] fixup! Windows nvcc bug workaround --- ggml-cuda.cu | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ggml-cuda.cu b/ggml-cuda.cu index 8e6ded09b..a62f26e1e 100644 --- a/ggml-cuda.cu +++ b/ggml-cuda.cu @@ -1516,7 +1516,7 @@ static void ggml_cuda_op(const ggml_tensor * src0, const ggml_tensor * src1, ggm // There is possibly a bug in the Windows nvcc compiler regarding instruction reordering or optimizing out local variables. // Removing the first assert or changing the order of the arguments causes the second assert to fail. // Removing both asserts results in i01_high becoming 0 which in turn results in garbage output. - // The root cause seems to be a problem with i0_offset_high becoming 0 when it should always be 1 (for single GPU). + // The root cause seems to be a problem with i0_offset_high becoming 0 when it should always be >0 (for single GPU). GGML_ASSERT(i01_low == 0 || g_device_count > 1); GGML_ASSERT(i01_high == ne01 || g_device_count > 1);