From 43812b36d763259216b288c62ea343e131510244 Mon Sep 17 00:00:00 2001
From: psychocrypt <psychocrypt@users.noreply.github.com>
Date: Thu, 23 Nov 2017 21:25:25 +0100
Subject: fix CUDA launch bounds usage

fix #191

lauch bounds must be placed before the return type but after the template paramater
---
 xmrstak/backend/nvidia/nvcc_code/cuda_core.cu | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'xmrstak/backend/nvidia/nvcc_code')

diff --git a/xmrstak/backend/nvidia/nvcc_code/cuda_core.cu b/xmrstak/backend/nvidia/nvcc_code/cuda_core.cu
index a92fa8c..dba6676 100644
--- a/xmrstak/backend/nvidia/nvcc_code/cuda_core.cu
+++ b/xmrstak/backend/nvidia/nvcc_code/cuda_core.cu
@@ -167,10 +167,10 @@ __forceinline__ __device__ uint32_t shuffle(volatile uint32_t* ptr,const uint32_
 #endif
 }
 
+template<size_t ITERATIONS, uint32_t THREAD_SHIFT, uint32_t MASK>
 #ifdef XMR_STAK_THREADS
 __launch_bounds__( XMR_STAK_THREADS * 4 )
 #endif
-template<size_t ITERATIONS, uint32_t THREAD_SHIFT, uint32_t MASK>
 __global__ void cryptonight_core_gpu_phase2( int threads, int bfactor, int partidx, uint32_t * d_long_state, uint32_t * d_ctx_a, uint32_t * d_ctx_b )
 {
 	__shared__ uint32_t sharedMemory[1024];
-- 
cgit v1.1