9 files changed, 86 insertions, 15 deletions
diff --git a/src/cuda-sim/cuda-sim.cc b/src/cuda-sim/cuda-sim.cc
index 9d62a59..c471a5c 100644
--- a/src/cuda-sim/cuda-sim.cc
+++ b/src/cuda-sim/cuda-sim.cc
@@ -83,7 +83,7 @@
 #include "ptx_sim.h"
 #include "../gpgpusim_entrypoint.h"
 
-int gpgpu_ptx_instruction_classification=0;
+int gpgpu_ptx_instruction_classification;
 void ** g_inst_classification_stat = NULL;
 void ** g_inst_op_classification_stat= NULL;
 int g_ptx_kernel_count = -1; // used for classification stat collection purposes 
@@ -1327,7 +1327,7 @@ void gpgpu_ptx_sim_memcpy_symbol(const char *hostVar, const void *src, size_t co
    fflush(stdout);
 }
 
-int g_ptx_sim_mode=0; // if non-zero run functional simulation only (i.e., no notion of a clock cycle)
+int g_ptx_sim_mode; // if non-zero run functional simulation only (i.e., no notion of a clock cycle)
 
 extern "C" int ptx_debug;
 
diff --git a/src/cuda-sim/cuda_device_printf.cc b/src/cuda-sim/cuda_device_printf.cc
index 197a5dd..9edeb5e 100644
--- a/src/cuda-sim/cuda_device_printf.cc
+++ b/src/cuda-sim/cuda_device_printf.cc
@@ -62,6 +62,7 @@
  * Vancouver, BC V6T 1Z4
  */
 
+#include "cuda_device_printf.h"
 #include "ptx_ir.h"
 
 void decode_space( memory_space_t &space, const ptx_thread_info *thread, const operand_info &op, memory_space *&mem, addr_t &addr);
diff --git a/src/cuda-sim/cuda_device_printf.h b/src/cuda-sim/cuda_device_printf.h
new file mode 100644
index 0000000..022b555
--- /dev/null
+++ b/src/cuda-sim/cuda_device_printf.h
@@ -0,0 +1,70 @@
+/* 
+ * Copyright (c) 2009 by Tor M. Aamodt, Ali Bakhoda, Joey Ting, Dan O'Connor, 
+ * Clive Lin, George L. Yuan, Wilson W. L. Fung and the 
+ * University of British Columbia
+ * Vancouver, BC  V6T 1Z4
+ * All Rights Reserved.
+ * 
+ * THIS IS A LEGAL DOCUMENT BY DOWNLOADING GPGPU-SIM, YOU ARE AGREEING TO THESE
+ * TERMS AND CONDITIONS.
+ * 
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNERS OR CONTRIBUTORS BE
+ * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+ * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+ * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+ * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+ * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+ * POSSIBILITY OF SUCH DAMAGE.
+ * 
+ * NOTE: The files libcuda/cuda_runtime_api.c and src/cuda-sim/cuda-math.h
+ * are derived from the CUDA Toolset available from http://www.nvidia.com/cuda
+ * (property of NVIDIA).  The files benchmarks/BlackScholes/ and 
+ * benchmarks/template/ are derived from the CUDA SDK available from 
+ * http://www.nvidia.com/cuda (also property of NVIDIA).  The files from 
+ * src/intersim/ are derived from Booksim (a simulator provided with the 
+ * textbook "Principles and Practices of Interconnection Networks" available 
+ * from http://cva.stanford.edu/books/ppin/). As such, those files are bound by 
+ * the corresponding legal terms and conditions set forth separately (original 
+ * copyright notices are left in files from these sources and where we have 
+ * modified a file our copyright notice appears before the original copyright 
+ * notice).  
+ * 
+ * Using this version of GPGPU-Sim requires a complete installation of CUDA 
+ * which is distributed seperately by NVIDIA under separate terms and 
+ * conditions.  To use this version of GPGPU-Sim with OpenCL requires a
+ * recent version of NVIDIA's drivers which support OpenCL.
+ * 
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ * 
+ * 1. Redistributions of source code must retain the above copyright notice,
+ * this list of conditions and the following disclaimer.
+ * 
+ * 2. Redistributions in binary form must reproduce the above copyright notice,
+ * this list of conditions and the following disclaimer in the documentation
+ * and/or other materials provided with the distribution.
+ * 
+ * 3. Neither the name of the University of British Columbia nor the names of
+ * its contributors may be used to endorse or promote products derived from
+ * this software without specific prior written permission.
+ * 
+ * 4. This version of GPGPU-SIM is distributed freely for non-commercial use only.  
+ *  
+ * 5. No nonprofit user may place any restrictions on the use of this software,
+ * including as modified by the user, by any other authorized user.
+ * 
+ * 6. GPGPU-SIM was developed primarily by Tor M. Aamodt, Wilson W. L. Fung, 
+ * Ali Bakhoda, George L. Yuan, at the University of British Columbia, 
+ * Vancouver, BC V6T 1Z4
+ */
+
+#ifndef CUDA_DEVICE_PRINTF_INCLUDED
+#define CUDA_DEVICE_PRINTF_INCLUDED
+
+void gpgpusim_cuda_vprintf(const class ptx_instruction * pI, const class ptx_thread_info * thread, const class function_info * target_func );
+
+#endif
diff --git a/src/cuda-sim/instructions.cc b/src/cuda-sim/instructions.cc
index 0924eab..47b3ebb 100644
--- a/src/cuda-sim/instructions.cc
+++ b/src/cuda-sim/instructions.cc
@@ -75,6 +75,9 @@
 #include "cuda-math.h"
 #include "../abstract_hardware_model.h"
 #include "ptx_loader.h"
+#include "cuda_device_printf.h"
+#include "../gpgpu-sim/gpu-sim.h"
+#include "../gpgpu-sim/shader.h"
 
 #include <stdarg.h>
 
@@ -678,11 +681,6 @@ void bra_impl( const ptx_instruction *pI, ptx_thread_info *thread )
 void brev_impl( const ptx_instruction *pI, ptx_thread_info *thread ) { inst_not_implemented(pI); }
 void brkpt_impl( const ptx_instruction *pI, ptx_thread_info *thread ) { inst_not_implemented(pI); }
 
-extern int gpgpu_simd_model;
-#define POST_DOMINATOR 1 /* must match enum value in shader.h */
-void get_pdom_stack_top_info( unsigned sid, unsigned tid, unsigned *npc, unsigned *rpc );
-void gpgpusim_cuda_vprintf(const ptx_instruction * pI, const ptx_thread_info * thread, const function_info * target_func ); 
-
 void call_impl( const ptx_instruction *pI, ptx_thread_info *thread ) 
 {
    static unsigned call_uid_next = 1;
diff --git a/src/cuda-sim/ptx-stats.cc b/src/cuda-sim/ptx-stats.cc
index 6e7e6ec..df95960 100644
--- a/src/cuda-sim/ptx-stats.cc
+++ b/src/cuda-sim/ptx-stats.cc
@@ -71,7 +71,7 @@
 extern function_info *g_func_info;
 
 // options
-int enable_ptx_file_line_stats;
+bool enable_ptx_file_line_stats;
 char * ptx_line_stats_filename = NULL;
 
 void ptx_file_line_stats_options(option_parser_t opp)
diff --git a/src/cuda-sim/ptx-stats.h b/src/cuda-sim/ptx-stats.h
index 200f7a5..004433b 100644
--- a/src/cuda-sim/ptx-stats.h
+++ b/src/cuda-sim/ptx-stats.h
@@ -63,7 +63,7 @@
 
 #include "../option_parser.h"
 
-extern int enable_ptx_file_line_stats;
+extern bool enable_ptx_file_line_stats;
 
 // set options
 void ptx_file_line_stats_options(option_parser_t opp);
diff --git a/src/cuda-sim/ptx_ir.h b/src/cuda-sim/ptx_ir.h
index ce68505..3a2a4e3 100644
--- a/src/cuda-sim/ptx_ir.h
+++ b/src/cuda-sim/ptx_ir.h
@@ -1272,6 +1272,8 @@ extern std::map<std::string,symbol_table*> g_sym_name_to_symbol_table;
 
 #define STATIC_ALLOC_LIMIT (GLOBAL_HEAP_START - (TOTAL_LOCAL_MEM+TOTAL_SHARED_MEM))
 
+extern bool g_keep_intermediate_files;
+
 void gpgpu_ptx_assemble( std::string kname, void *kinfo );
 #include "../option_parser.h"
 void ptx_reg_options(option_parser_t opp);
diff --git a/src/cuda-sim/ptx_loader.cc b/src/cuda-sim/ptx_loader.cc
index 713bdb3..783ce91 100644
--- a/src/cuda-sim/ptx_loader.cc
+++ b/src/cuda-sim/ptx_loader.cc
@@ -299,13 +299,17 @@ void gpgpu_ptx_sim_add_ptxstring( const char *ptx_string, const char *sourcefnam
     }
 }
 
-static int g_save_embedded_ptx = 0;
+static bool g_save_embedded_ptx;
+bool g_keep_intermediate_files;
 
 void ptx_reg_options(option_parser_t opp)
 {
    option_parser_register(opp, "-save_embedded_ptx", OPT_BOOL, &g_save_embedded_ptx, 
                 "saves ptx files embedded in binary as <n>.ptx",
                 "0");
+   option_parser_register(opp, "-keep", OPT_BOOL, &g_keep_intermediate_files, 
+                "keep intermediate files created by GPGPU-Sim when interfacing with external programs",
+                "0");
 }
 
 void print_ptx_file( const char *p, unsigned source_num, const char *filename )
diff --git a/src/cuda-sim/ptx_sim.cc b/src/cuda-sim/ptx_sim.cc
index 2156bd6..b433fb3 100644
--- a/src/cuda-sim/ptx_sim.cc
+++ b/src/cuda-sim/ptx_sim.cc
@@ -65,6 +65,7 @@
 #include <string>
 #include "ptx_ir.h"
 #include "../gpgpu-sim/gpu-sim.h"
+#include "../gpgpu-sim/shader.h"
 
 void feature_not_implemented( const char *f );
 
@@ -358,11 +359,6 @@ void ptx_thread_info::callstack_push( unsigned pc, unsigned rpc, const symbol *r
    m_local_mem_stack_pointer += m_func_info->local_mem_framesize(); 
 }
 
-#define POST_DOMINATOR 1 /* must match definition in shader.h */
-extern int gpgpu_simd_model;
-extern ptx_reg_t get_operand_value( const symbol *reg );
-extern void set_operand_value( const symbol *dst, const ptx_reg_t &data );
-
 bool ptx_thread_info::callstack_pop()
 {
    const symbol *rv_src = m_callstack.back().m_return_var_src;