- BlackScholes passing on CUDA 3.1

- Added fma operation by simply calling mad, however in reality fma is supposed to
  be more accurate than mad.  For now, properly emulating fma (to get the extra
  precision) isn't a priority.

[git-p4: depot-paths = "//depot/gpgpu_sim_research/fermi/distribution/": change = 6839]

1 files changed, 12 insertions, 0 deletions

diff --git a/src/cuda-sim/instructions.cc b/src/cuda-sim/instructions.cc
index 4479cbe..226e965 100644
--- a/src/cuda-sim/instructions.cc
+++ b/src/cuda-sim/instructions.cc
@@ -1386,6 +1386,13 @@ void exit_impl( const ptx_instruction *pI, ptx_thread_info *thread )
    thread->set_done();
 }
 
+void mad_def( const ptx_instruction *pI, ptx_thread_info *thread );
+
+void fma_impl( const ptx_instruction *pI, ptx_thread_info *thread ) 
+{
+   mad_def(pI,thread);
+}
+
 void ld_impl( const ptx_instruction *pI, ptx_thread_info *thread ) 
 { 
    const operand_info &dst = pI->dst();
@@ -1502,6 +1509,11 @@ void mad24_impl( const ptx_instruction *pI, ptx_thread_info *thread )
 }
 
 void mad_impl( const ptx_instruction *pI, ptx_thread_info *thread ) 
+{
+   mad_def(pI,thread);
+}
+
+void mad_def( const ptx_instruction *pI, ptx_thread_info *thread ) 
 { 
    const operand_info &dst  = pI->dst();
    const operand_info &src1 = pI->src1();