i#1729 offline traces: generate unoptimized traces

Adds two new options for offline trace generation in drcachesim: -offline
and -outdir.

Updates the launcher to skip the simulator launch when -offline is
requested.

Updates the tracer to write to a file instead of a pipe when -offline is
set.  The trace entry format is still the same: it will be modified in the
future to include timestamps and to shrink its size, with a postprocessor
component filling in statically-identifiable information.

Adds packing to trace_entry_t to ensure it can be used as a cross-platform
persistent representation.

A test will be added once file reading is in place.

Review-URL: https://codereview.appspot.com/310890043

clients/drcachesim/common/options.cpp

@@ -1,5 +1,5 @@
 /* **********************************************************
- * Copyright (c) 2015 Google, Inc.  All rights reserved.
+ * Copyright (c) 2015-2016 Google, Inc.  All rights reserved.
  * **********************************************************/
 
 /*
@@ -36,12 +36,24 @@
 #include "droption.h"
 #include "options.h"
 
+droption_t<bool> op_offline
+(DROPTION_SCOPE_ALL, "offline", false, "Store trace files for offline analysis",
+ "By default, traces are processed online, sent over a pipe to a simulator.  "
+ "If this option is enabled, trace data is instead written to files in -outdir "
+ "for later offline analysis.  No simulator is executed.");
+
 droption_t<std::string> op_ipc_name
 (DROPTION_SCOPE_ALL, "ipc_name", "drcachesimpipe", "Base name of named pipe",
- "Specifies the base name of the named pipe used to communicate between the target "
+ "For online tracing and simulation (the default, unless -offline is requested), "
+ "specifies the base name of the named pipe used to communicate between the target "
  "application processes and the caching device simulator.  A unique name must be chosen "
  "for each instance of the simulator being run at any one time.");
 
+droption_t<std::string> op_outdir
+(DROPTION_SCOPE_ALL, "outdir", ".", "Target directory for offline trace files",
+ "For the offline analysis mode (when -offline is requested), specifies the path "
+ "to a directory where per-thread trace files will be written.");
+
 droption_t<unsigned int> op_num_cores
 (DROPTION_SCOPE_FRONTEND, "cores", 4, "Number of cores",
  "Specifies the number of cores to simulate.");

clients/drcachesim/common/options.h

@@ -1,5 +1,5 @@
 /* **********************************************************
- * Copyright (c) 2015 Google, Inc.  All rights reserved.
+ * Copyright (c) 2015-2016 Google, Inc.  All rights reserved.
  * **********************************************************/
 
 /*
@@ -45,7 +45,9 @@
 #include <string>
 #include "droption.h"
 
+extern droption_t<bool> op_offline;
 extern droption_t<std::string> op_ipc_name;
+extern droption_t<std::string> op_outdir;
 extern droption_t<unsigned int> op_num_cores;
 extern droption_t<unsigned int> op_line_size;
 extern droption_t<bytesize_t> op_L1I_size;

clients/drcachesim/common/trace_entry.h

@@ -1,5 +1,5 @@
 /* **********************************************************
- * Copyright (c) 2015 Google, Inc.  All rights reserved.
+ * Copyright (c) 2015-2016 Google, Inc.  All rights reserved.
  * **********************************************************/
 
 /*
@@ -44,6 +44,7 @@
 #define _TRACE_ENTRY_H_ 1
 
 #include <stdint.h>
+#include "utils.h"
 
 typedef uintptr_t addr_t;
 
@@ -116,7 +117,8 @@ extern const char * const trace_type_names[];
 // - a flush request
 // - a prefetch request
 // - a thread/process
-typedef struct _trace_entry_t {
+START_PACKED_STRUCTURE
+struct _trace_entry_t {
     unsigned short type; // 2 bytes: trace_type_t
     // 2 bytes: mem ref size, instr length, or num of instrs for instr bundle
     unsigned short size;
@@ -125,7 +127,8 @@ typedef struct _trace_entry_t {
         // The length of each instr in the instr bundle
         unsigned char length[sizeof(addr_t)];
     };
-} trace_entry_t;
+} END_PACKED_STRUCTURE;
+typedef struct _trace_entry_t trace_entry_t;
 
 static inline bool
 type_is_prefetch(unsigned short type)

clients/drcachesim/common/utils.h

@@ -1,5 +1,5 @@
 /* **********************************************************
- * Copyright (c) 2015 Google, Inc.  All rights reserved.
+ * Copyright (c) 2015-2016 Google, Inc.  All rights reserved.
  * **********************************************************/
 
 /*
@@ -48,6 +48,25 @@
 #define BUFFER_LAST_ELEMENT(buf)    buf[BUFFER_SIZE_ELEMENTS(buf) - 1]
 #define NULL_TERMINATE_BUFFER(buf)  BUFFER_LAST_ELEMENT(buf) = 0
 
+#define BOOLS_MATCH(b1, b2) (!!(b1) == !!(b2))
+
+#ifdef WINDOWS
+/* Use special C99 operator _Pragma to generate a pragma from a macro */
+# if _MSC_VER <= 1200
+#  define ACTUAL_PRAGMA(p) _Pragma ( #p )
+# else
+#  define ACTUAL_PRAGMA(p) __pragma ( p )
+# endif
+/* Usage: if planning to typedef, that must be done separately, as MSVC will
+ * not take _pragma after typedef.
+ */
+# define START_PACKED_STRUCTURE ACTUAL_PRAGMA( pack(push,1) )
+# define END_PACKED_STRUCTURE ACTUAL_PRAGMA( pack(pop) )
+#else
+# define START_PACKED_STRUCTURE /* nothing */
+# define END_PACKED_STRUCTURE __attribute__ ((__packed__))
+#endif
+
 static inline int
 compute_log2(int value)
 {

clients/drcachesim/launcher.cpp

@@ -1,5 +1,5 @@
 /* **********************************************************
- * Copyright (c) 2015 Google, Inc.  All rights reserved.
+ * Copyright (c) 2015-2016 Google, Inc.  All rights reserved.
  * **********************************************************/
 
 /*
@@ -81,6 +81,13 @@ file_is_readable(const char *path)
     return (drfront_access(path, DRFRONT_READ, &ret) == DRFRONT_SUCCESS && ret);
 }
 
+static bool
+file_is_writable(const char *path)
+{
+    bool ret = false;
+    return (drfront_access(path, DRFRONT_WRITE, &ret) == DRFRONT_SUCCESS && ret);
+}
+
 static void
 get_full_path(const char *app, char *buf, size_t buflen/*# elements*/)
 {
@@ -157,6 +164,9 @@ _tmain(int argc, const TCHAR *targv[])
     bool is64, is32;
     analyzer_t *analyzer = NULL;
     std::string tracer_ops;
+#ifdef UNIX
+    pid_t child;
+#endif
 
 #if defined(WINDOWS) && !defined(_UNICODE)
 # error _UNICODE must be defined
@@ -222,20 +232,29 @@ _tmain(int argc, const TCHAR *targv[])
         assert(false); // won't get here
     }
 
-    // declare the analyzer based on its type
-    if (op_simulator_type.get_value() == CPU_CACHE)
-        analyzer = new cache_simulator_t;
-    else if (op_simulator_type.get_value() == TLB)
-        analyzer = new tlb_simulator_t;
-    else {
-        ERROR("Usage error: unsupported analyzer type. "
-              "Please choose " CPU_CACHE" or " TLB".\n");
-        return false;
-    }
+    if (op_offline.get_value()) {
+        // Initial sanity check: may still be unwritable by this user, but this
+        // serves as at least an existence check.
+        if (!file_is_writable(op_outdir.get_value().c_str())) {
+            FATAL_ERROR("invalid -outdir %s", op_outdir.get_value().c_str());
+            assert(false); // won't get here
+        }
+    } else {
+        // declare the analyzer based on its type
+        if (op_simulator_type.get_value() == CPU_CACHE)
+            analyzer = new cache_simulator_t;
+        else if (op_simulator_type.get_value() == TLB)
+            analyzer = new tlb_simulator_t;
+        else {
+            ERROR("Usage error: unsupported analyzer type. "
+                  "Please choose " CPU_CACHE" or " TLB".\n");
+            return false;
+        }
 
-    if (!analyzer->init()) {
-        FATAL_ERROR("failed to initialize analyzer");
-        assert(false); // won't get here
+        if (!analyzer->init()) {
+            FATAL_ERROR("failed to initialize analyzer");
+            assert(false); // won't get here
+        }
     }
 
     tracer_ops = op_tracer_ops.get_value();
@@ -245,12 +264,15 @@ _tmain(int argc, const TCHAR *targv[])
     NOTIFY(1, "INFO", "DynamoRIO configuration directory is %s", buf);
 
 #ifdef UNIX
-    pid_t child = fork();
+    if (op_offline.get_value())
+        child = 0;
+    else
+        child = fork();
     if (child < 0) {
         FATAL_ERROR("failed to fork");
         assert(false); // won't get here
     } else if (child == 0) {
-        /* child */
+        /* child, or offline where we exec this process */
         if (!configure_application(app_name, app_argv, tracer_ops, &inject_data) ||
             !dr_inject_process_inject(inject_data, false/*!force*/, NULL)) {
             FATAL_ERROR("unable to inject");
@@ -268,9 +290,11 @@ _tmain(int argc, const TCHAR *targv[])
     dr_inject_process_run(inject_data);
 #endif
 
-    if (!analyzer->run()) {
-        FATAL_ERROR("failed to run analyzer");
-        assert(false); // won't get here
+    if (!op_offline.get_value()) {
+        if (!analyzer->run()) {
+            FATAL_ERROR("failed to run analyzer");
+            assert(false); // won't get here
+        }
     }
 
 #ifdef WINDOWS