From bdb302be663e50cb391e0db79c3baf14b59b04da Mon Sep 17 00:00:00 2001
From: TeseoSanJose <miguelteseomingot@gmail.com>
Date: Tue, 11 Feb 2025 09:48:07 +0000
Subject: [PATCH] Solves Hadrons compilation error adding extra templates

---
 Grid/threads/Accelerator.h | 17 +++++++++++------
 1 file changed, 11 insertions(+), 6 deletions(-)

diff --git a/Grid/threads/Accelerator.h b/Grid/threads/Accelerator.h
index c28ca201..9ae282f2 100644
--- a/Grid/threads/Accelerator.h
+++ b/Grid/threads/Accelerator.h
@@ -215,7 +215,7 @@ inline void *acceleratorAllocHost(size_t bytes)
   auto err = cudaMallocHost((void **)&ptr,bytes);
   if( err != cudaSuccess ) {
     ptr = (void *) NULL;
-    printf(" cudaMallocHost failed for %d %s \n",bytes,cudaGetErrorString(err));
+    printf(" cudaMallocHost failed for %ld %s \n",bytes,cudaGetErrorString(err));
     assert(0);
   }
   return ptr;
@@ -226,7 +226,7 @@ inline void *acceleratorAllocShared(size_t bytes)
   auto err = cudaMallocManaged((void **)&ptr,bytes);
   if( err != cudaSuccess ) {
     ptr = (void *) NULL;
-    printf(" cudaMallocManaged failed for %d %s \n",bytes,cudaGetErrorString(err));
+    printf(" cudaMallocManaged failed for %ld %s \n",bytes,cudaGetErrorString(err));
     assert(0);
   }
   return ptr;
@@ -237,7 +237,7 @@ inline void *acceleratorAllocDevice(size_t bytes)
   auto err = cudaMalloc((void **)&ptr,bytes);
   if( err != cudaSuccess ) {
     ptr = (void *) NULL;
-    printf(" cudaMalloc failed for %d %s \n",bytes,cudaGetErrorString(err));
+    printf(" cudaMalloc failed for %ld %s \n",bytes,cudaGetErrorString(err));
   }
   return ptr;
 };
@@ -245,7 +245,8 @@ inline void *acceleratorAllocDevice(size_t bytes)
 inline void acceleratorFreeShared(void *ptr){ cudaFree(ptr);};
 inline void acceleratorFreeDevice(void *ptr){ cudaFree(ptr);};
 inline void acceleratorFreeHost(void *ptr){ cudaFree(ptr);};
-inline void acceleratorCopyToDevice(void *from,void *to,size_t bytes)  { cudaMemcpy(to,from,bytes, cudaMemcpyHostToDevice);}
+inline void acceleratorCopyToDevice(void *from, void *to, size_t bytes)  { cudaMemcpy(to,from,bytes, cudaMemcpyHostToDevice);}
+inline void acceleratorCopyToDevice(const void *from, void *to, size_t bytes) {cudaMemcpy(to,from,bytes, cudaMemcpyHostToDevice);}
 inline void acceleratorCopyFromDevice(void *from,void *to,size_t bytes){ cudaMemcpy(to,from,bytes, cudaMemcpyDeviceToHost);}
 inline void acceleratorCopyToDeviceAsync(void *from, void *to, size_t bytes, cudaStream_t stream = copyStream) { cudaMemcpyAsync(to,from,bytes, cudaMemcpyHostToDevice, stream);}
 inline void acceleratorCopyFromDeviceAsync(void *from, void *to, size_t bytes, cudaStream_t stream = copyStream) { cudaMemcpyAsync(to,from,bytes, cudaMemcpyDeviceToHost, stream);}
@@ -649,7 +650,7 @@ accelerator_inline void acceleratorFence(void)
   return;
 }
 
-inline void acceleratorCopyDeviceToDevice(void *from,void *to,size_t bytes)
+inline void acceleratorCopyDeviceToDevice(void *from,void *to, size_t bytes)
 {
   acceleratorCopyDeviceToDeviceAsynch(from,to,bytes);
   acceleratorCopySynchronise();
@@ -657,7 +658,11 @@ inline void acceleratorCopyDeviceToDevice(void *from,void *to,size_t bytes)
 
 template<class T> void acceleratorPut(T& dev,T&host)
 {
-  acceleratorCopyToDevice(&host,&dev,sizeof(T));
+  acceleratorCopyToDevice(&host, &dev, sizeof(T));
+}
+template<class T> void acceleratorPut(T& dev, const T& host)
+{
+  acceleratorCopyToDevice(&host, &dev, sizeof(T));
 }
 template<class T> T acceleratorGet(T& dev)
 {