1. enable L2 cache as a texture cache (also some bug fixes for L2 as regular cache)

2. update gpgpusim.config for Quadro to use L1 cache geometry from Henry's ISPASS paper
3. minor edit to CUDA api : add notion of fat_cubin_handle (currently not used for anything)
4. minor edits to deadlock detection message (more accurate reporting of source of deadlock)
5. other minor edits

[git-p4: depot-paths = "//depot/gpgpu_sim_research/fermi/distribution/": change = 7809]

1 files changed, 1 insertions, 5 deletions

diff --git a/src/cuda-sim/cuda-sim.cc b/src/cuda-sim/cuda-sim.cc
index de4445a..afadbe4 100644
--- a/src/cuda-sim/cuda-sim.cc
+++ b/src/cuda-sim/cuda-sim.cc
@@ -1396,7 +1396,7 @@ void print_splash()
    }
 }
 
-void gpgpu_ptx_sim_register_kernel(const char *hostFun, const char *deviceFun)
+void gpgpu_ptx_sim_register_kernel(void **fatCubinHandle,const char *hostFun, const char *deviceFun)
 {
    const void* key=hostFun;
    print_splash();
@@ -1428,16 +1428,12 @@ void gpgpu_ptx_sim_register_const_variable(void *hostVar, const char *deviceName
 {
    printf("GPGPU-Sim PTX registering constant %s (%zu bytes) to name mapping\n", deviceName, size );
    g_const_name_lookup[hostVar] = deviceName;
-   //assert( g_current_symbol_table != NULL );
-   //g_sym_name_to_symbol_table[deviceName] = g_current_symbol_table;
 }
 
 void gpgpu_ptx_sim_register_global_variable(void *hostVar, const char *deviceName, size_t size )
 {
    printf("GPGPU-Sim PTX registering global %s hostVar to name mapping\n", deviceName );
    g_global_name_lookup[hostVar] = deviceName;
-   //assert( g_current_symbol_table != NULL );
-   //g_sym_name_to_symbol_table[deviceName] = g_current_symbol_table;
 }
 
 void gpgpu_ptx_sim_memcpy_symbol(const char *hostVar, const void *src, size_t count, size_t offset, int to )