- reduce 3.1 sdk regression test runtimes (as per 2.3 sdk versions)

- add ldu support (just use ld for now)
- get rid of asserts for .const[n] 
- status:
	cuda sdk 3.1 dct8x8, eigenvalues, MersenneTwister, reduction, scan still running (too slow)
	ocl sdk 3.1 oclQuasirandomGenerator still running (too slow)
	ocl sdk 3.1 oclReduction fail comparison (cuda 3.1)
	ocl sdk 3.1 oclMatrixMul exits mid computation
	ocl sdk 3.1 oclTranspose hits "clCreateBuffer - buffer already created for this host variable"
	ocl sdk 3.1 oclSortingNetworks fails w/ message about min workgroup size
	remaining 30 tests pass

[git-p4: depot-paths = "//depot/gpgpu_sim_research/fermi/distribution/": change = 6927]

1 files changed, 10 insertions, 3 deletions

diff --git a/src/cuda-sim/instructions.cc b/src/cuda-sim/instructions.cc
index a19f1a8..3ce0a2f 100644
--- a/src/cuda-sim/instructions.cc
+++ b/src/cuda-sim/instructions.cc
@@ -1452,7 +1452,7 @@ void decode_space( memory_space_t &space, const ptx_thread_info *thread, const o
    case surf_space:   mem = g_surf_mem; break; 
    case param_space_kernel:  mem = g_param_mem; break;
    case shared_space:  mem = thread->m_shared_mem; break; 
-   case const_space:  assert(space.get_bank()==0); mem = g_global_mem; break;
+   case const_space:  mem = g_global_mem; break;
    case generic_space:
       if( thread->get_ptx_version().ver() >= 2.0 ) {
          // convert generic address to memory space address
@@ -1474,7 +1474,7 @@ void decode_space( memory_space_t &space, const ptx_thread_info *thread, const o
    }
 }
 
-void ld_impl( const ptx_instruction *pI, ptx_thread_info *thread ) 
+void ld_exec( const ptx_instruction *pI, ptx_thread_info *thread ) 
 { 
    const operand_info &dst = pI->dst();
    const operand_info &src1 = pI->src1();
@@ -1515,7 +1515,14 @@ void ld_impl( const ptx_instruction *pI, ptx_thread_info *thread )
    thread->m_last_memory_space = space; 
 }
 
-void ldu_impl( const ptx_instruction *pI, ptx_thread_info *thread ) { inst_not_implemented(pI); }
+void ld_impl( const ptx_instruction *pI, ptx_thread_info *thread ) 
+{
+   ld_exec(pI,thread);
+}
+void ldu_impl( const ptx_instruction *pI, ptx_thread_info *thread ) 
+{ 
+   ld_exec(pI,thread);
+}
 
 void lg2_impl( const ptx_instruction *pI, ptx_thread_info *thread ) 
 {