Commit 8dbe1eb4 authored by Tomáš Oberhuber's avatar Tomáš Oberhuber
Browse files

Fixing CUDA reduction kernels for the CUDA 3.0.

Fixining yywrap() in debug scanner.
parent b4d79182
Loading
Loading
Loading
Loading
+12 −6
Original line number Diff line number Diff line
@@ -177,7 +177,8 @@ __global__ void tnlCUDAReductionKernel( const int size,
                                        T* deviceOutput,
                                        T* dbg_array1 = 0 )
{
   extern __shared__ __align__ ( 8 ) T sdata[];
   extern __shared__ __align__ ( 8 ) char __sdata[];
   T* sdata = reinterpret_cast< T* >( __sdata );

   /***
    * Get thread id (tid) and global thread id (gid).
@@ -513,7 +514,8 @@ __global__ void tnlCUDASimpleReductionKernel5( const int size,
                                               T* deviceOutput,
                                               T* dbg_array1 = 0 )
{
   extern __shared__ T sdata[];
   extern __shared__ __align__ ( 8 ) char __sdata[];
   T* sdata = reinterpret_cast< T* >( __sdata );

   /***
    * Get thread id (tid) and global thread id (gid).
@@ -732,7 +734,8 @@ __global__ void tnlCUDASimpleReductionKernel4( const int size,
	                                       T* deviceOutput,
	                   	               T* dbg_array1 = 0  )
{
   extern __shared__ T sdata[];
   extern __shared__ __align__ ( 8 ) char __sdata[];
   T* sdata = reinterpret_cast< T* >( __sdata );

   /***
    * Get thread id (tid) and global thread id (gid).
@@ -965,7 +968,8 @@ __global__ void tnlCUDASimpleReductionKernel3( const int size,
                                               const T* deviceInput,
		                               T* deviceOutput )
{
   extern __shared__ T sdata[];
   extern __shared__ __align__ ( 8 ) char __sdata[];
   T* sdata = reinterpret_cast< T* >( __sdata );

   /***
    * Get thread id (tid) and global thread id (gid).
@@ -1157,7 +1161,8 @@ __global__ void tnlCUDASimpleReductionKernel2( const int size,
		                               const T* deviceInput,
		                               T* deviceOutput )
{
   extern __shared__ T sdata[];
   extern __shared__ __align__ ( 8 ) char __sdata[];
   T* sdata = reinterpret_cast< T* >( __sdata );

   /***
    * Get thread id (tid) and global thread id (gid).
@@ -1335,7 +1340,8 @@ __global__ void tnlCUDASimpleReductionKernel1( const int size,
		                               const T* deviceInput,
		                               T* deviceOutput )
{
   extern __shared__ T sdata[];
   extern __shared__ __align__ ( 8 ) char __sdata[];
   T* sdata = reinterpret_cast< T* >( __sdata );

   /***
    * Get thread id (tid) and global thread id (gid).
+1 −1
Original line number Diff line number Diff line
// Generated by Bisonc++ V2.4.7 on Sat, 19 Dec 2009 18:04:01 +0100
// Generated by Bisonc++ V2.4.7 on Thu, 27 May 2010 13:23:09 +0200

// $insert class.ih
#include "tnlDebugParser.ih"
+1 −1
Original line number Diff line number Diff line
// Generated by Bisonc++ V2.4.7 on Sat, 19 Dec 2009 18:04:01 +0100
// Generated by Bisonc++ V2.4.7 on Thu, 27 May 2010 13:23:09 +0200

#ifndef tnlDebugParserBase_h_included
#define tnlDebugParserBase_h_included
+4 −4
Original line number Diff line number Diff line
@@ -1944,13 +1944,13 @@ void tnlDebugfree (void * ptr )



int tnlDebugFlexLexer::yywrap()
/*int tnlDebugFlexLexer::yywrap()
{
   return 1;
}
}*/

/*int yywrap()
{
extern int yywrap();
/*{
   return 1;
}*/
+4 −4
Original line number Diff line number Diff line
@@ -41,12 +41,12 @@ default-debug { //cout << "DEFAULT" << endl;
.                                             { return yytext[ 0 ]; }
%%

int tnlDebugFlexLexer::yywrap()
/*int tnlDebugFlexLexer::yywrap()
{
   return 1;
}
}*/

/*int yywrap()
{
extern int yywrap();
/*{
   return 1;
}*/