- Adding support for cudaFuncSetCacheConfig API, that allows changing the

      L1 Cache and Shared Memory configurations across kernels. The support
      enable the user to specify two more configurations (Preferred L1) or
      (Preferred Shared Memory) besides the default config. If the
      cudaFuncSetCacheConfig API is used to set the cache configuration
      of a specific kernel to either of these configuration (cudaFuncCachePreferShared,
      cudaFuncCachePreferL1), the simulator will change the cache configuration
      at kernel launch accordingly, if there is no alternative configurations
      provided to the simulator it will use the default configurations with a
      warning message display

[git-p4: depot-paths = "//depot/gpgpu_sim_research/fermi/distribution/": change = 15816]

1 files changed, 7 insertions, 1 deletions

diff --git a/src/abstract_hardware_model.h b/src/abstract_hardware_model.h
index de6eef8..5d74651 100644
--- a/src/abstract_hardware_model.h
+++ b/src/abstract_hardware_model.h
@@ -226,6 +226,9 @@ struct core_config {
         m_valid = false; 
         num_shmem_bank=16; 
         shmem_limited_broadcast = false; 
+        gpgpu_shmem_sizeDefault=(unsigned)-1;
+        gpgpu_shmem_sizePrefL1=(unsigned)-1;
+        gpgpu_shmem_sizePrefShared=(unsigned)-1;
     }
     virtual void init() = 0;
 
@@ -244,7 +247,10 @@ struct core_config {
         return ((addr/WORD_SIZE) % num_shmem_bank);
     }
     unsigned mem_warp_parts;  
-    unsigned gpgpu_shmem_size;
+    mutable unsigned gpgpu_shmem_size;
+    unsigned gpgpu_shmem_sizeDefault;
+    unsigned gpgpu_shmem_sizePrefL1;
+    unsigned gpgpu_shmem_sizePrefShared;
 
     // texture and constant cache line sizes (used to determine number of memory accesses)
     unsigned gpgpu_cache_texl1_linesize;