PR11246 cont'd: Separate script/stapconf caching

I'm separating the caching and creation logic for stapconf, so it can be
conditionalized on s.use_cache instead of s.use_script_cache.

* session.h (systemtap_session): Store base_hash for better reuse.
* hash.cxx (get_base_hash): Get the base from the session, or build it.
  (find_hash): Split into separate script/stapconf versions.
* cache.cxx (add_to_cache, get_from_cache): Ditto.
* main.cxx (main): Adapt caller, and delay stapconf until pass-4.

7 files changed, 69 insertions, 42 deletions

diff --git a/cache.cxx b/cache.cxx
index 1b083d69..6b51f21c 100644
--- a/cache.cxx
+++ b/cache.cxx
@@ -46,14 +46,10 @@ static void clean_cache(systemtap_session& s);
 
 
 void
-add_to_cache(systemtap_session& s)
+add_stapconf_to_cache(systemtap_session& s)
 {
   bool verbose = s.verbose > 1;
 
-  // PR10543: clean the cache *before* we try putting something new into it.
-  // We don't want to risk having the brand new contents being erased again.
-  clean_cache(s);
-
   string stapconf_src_path = s.tmpdir + "/" + s.stapconf_name;
   if (!copy_file(stapconf_src_path, s.stapconf_path, verbose))
     {
@@ -63,6 +59,17 @@ add_to_cache(systemtap_session& s)
       // s.use_script_cache = false;
       // return;
     }
+}
+
+
+void
+add_script_to_cache(systemtap_session& s)
+{
+  bool verbose = s.verbose > 1;
+
+  // PR10543: clean the cache *before* we try putting something new into it.
+  // We don't want to risk having the brand new contents being erased again.
+  clean_cache(s);
 
   string module_src_path = s.tmpdir + "/" + s.module_name + ".ko";
   STAP_PROBE2(stap, cache__add__module, module_src_path.c_str(), s.hash_path.c_str());
@@ -92,16 +99,10 @@ add_to_cache(systemtap_session& s)
 
 
 bool
-get_from_cache(systemtap_session& s)
+get_stapconf_from_cache(systemtap_session& s)
 {
   string stapconf_dest_path = s.tmpdir + "/" + s.stapconf_name;
-  string module_dest_path = s.tmpdir + "/" + s.module_name + ".ko";
-  string c_src_path = s.hash_path;
-  int fd_stapconf, fd_module, fd_c;
-
-  if (c_src_path.rfind(".ko") == (c_src_path.size() - 3))
-    c_src_path.resize(c_src_path.size() - 3);
-  c_src_path += ".c";
+  int fd_stapconf;
 
   // See if stapconf exists
   fd_stapconf = open(s.stapconf_path.c_str(), O_RDONLY);
@@ -122,7 +123,22 @@ get_from_cache(systemtap_session& s)
   close(fd_stapconf);
 
   if (s.verbose > 1)
-    clog << "Pass 3: using cached " << s.stapconf_path << endl;
+    clog << "Pass 4: using cached " << s.stapconf_path << endl;
+
+  return true;
+}
+
+
+bool
+get_script_from_cache(systemtap_session& s)
+{
+  string module_dest_path = s.tmpdir + "/" + s.module_name + ".ko";
+  string c_src_path = s.hash_path;
+  int fd_module, fd_c;
+
+  if (c_src_path.rfind(".ko") == (c_src_path.size() - 3))
+    c_src_path.resize(c_src_path.size() - 3);
+  c_src_path += ".c";
 
   // See if module exists
   fd_module = open(s.hash_path.c_str(), O_RDONLY);
diff --git a/cache.h b/cache.h
index 89f44a43..548fa8c8 100644
--- a/cache.h
+++ b/cache.h
@@ -1,4 +1,7 @@
-void add_to_cache(systemtap_session& s);
-bool get_from_cache(systemtap_session& s);
+void add_script_to_cache(systemtap_session& s);
+bool get_script_from_cache(systemtap_session& s);
+
+void add_stapconf_to_cache(systemtap_session& s);
+bool get_stapconf_from_cache(systemtap_session& s);
 
 /* vim: set sw=2 ts=8 cino=>4,n-2,{2,^-2,t0,(0,u0,w1,M1 : */
diff --git a/elaborate.cxx b/elaborate.cxx
index f442aceb..a2ab8522 100644
--- a/elaborate.cxx
+++ b/elaborate.cxx
@@ -1516,6 +1516,7 @@ semantic_pass (systemtap_session& s)
 
 systemtap_session::systemtap_session ():
   // NB: pointer members must be manually initialized!
+  base_hash(0),
   pattern_root(new match_node),
   user_file (0),
   be_derived_probes(0),
diff --git a/hash.cxx b/hash.cxx
index 36a71a42..6c37dbc7 100644
--- a/hash.cxx
+++ b/hash.cxx
@@ -104,9 +104,15 @@ void create_hash_log(const string &type_str, const string &parms, const string &
   log_file.close();
 }
 
-static void
-get_base_hash (systemtap_session& s, hash& h)
+static const hash&
+get_base_hash (systemtap_session& s)
 {
+  if (s.base_hash)
+    return *s.base_hash;
+
+  s.base_hash = new hash();
+  hash& h = *s.base_hash;
+
   // Hash kernel release and arch.
   h.add(s.kernel_release);
   h.add(s.kernel_build_tree);
@@ -141,6 +147,8 @@ get_base_hash (systemtap_session& s, hash& h)
   // know exactly where we're getting run from, we'll use
   // /proc/self/exe.
   h.add_file("/proc/self/exe");
+
+  return h;
 }
 
 
@@ -178,10 +186,10 @@ create_hashdir (systemtap_session& s, const string& result, string& hashdir)
 }
 
 
-static void
-find_script_hash (systemtap_session& s, const string& script, const hash &base)
+void
+find_script_hash (systemtap_session& s, const string& script)
 {
-  hash h(base);
+  hash h(get_base_hash(s));
   struct stat st;
 
   // Hash getuid.  This really shouldn't be necessary (since who you
@@ -260,10 +268,10 @@ find_script_hash (systemtap_session& s, const string& script, const hash &base)
 }
 
 
-static void
-find_stapconf_hash (systemtap_session& s, const hash& base)
+void
+find_stapconf_hash (systemtap_session& s)
 {
-  hash h(base);
+  hash h(get_base_hash(s));
 
   // Add any custom kbuild flags
   for (unsigned i = 0; i < s.kbuildflags.size(); i++)
@@ -282,21 +290,10 @@ find_stapconf_hash (systemtap_session& s, const hash& base)
 }
 
 
-void
-find_hash (systemtap_session& s, const string& script)
-{
-  hash base;
-  get_base_hash(s, base);
-  find_stapconf_hash(s, base);
-  find_script_hash(s, script, base);
-}
-
-
 string
 find_tracequery_hash (systemtap_session& s, const vector<string>& headers)
 {
-  hash h;
-  get_base_hash(s, h);
+  hash h(get_base_hash(s));
 
   // Add the tracepoint headers to the computed hash
   for (size_t i = 0; i < headers.size(); ++i)
@@ -321,8 +318,7 @@ find_tracequery_hash (systemtap_session& s, const vector<string>& headers)
 string
 find_typequery_hash (systemtap_session& s, const string& name)
 {
-  hash h;
-  get_base_hash(s, h);
+  hash h(get_base_hash(s));
 
   // Add the typequery name to distinguish the hash
   h.add(name);
diff --git a/hash.h b/hash.h
index 0c1a745f..bbe4e89c 100644
--- a/hash.h
+++ b/hash.h
@@ -41,7 +41,8 @@ public:
   std::string get_parms();
 };
 
-void find_hash (systemtap_session& s, const std::string& script);
+void find_script_hash (systemtap_session& s, const std::string& script);
+void find_stapconf_hash (systemtap_session& s);
 std::string find_tracequery_hash (systemtap_session& s,
                                   const std::vector<std::string>& headers);
 std::string find_typequery_hash (systemtap_session& s, const std::string& name);
diff --git a/main.cxx b/main.cxx
index 04431e72..6d3f649b 100644
--- a/main.cxx
+++ b/main.cxx
@@ -1300,10 +1300,10 @@ main (int argc, char * const argv [])
       }
 
       // Generate hash
-      find_hash (s, o.str());
+      find_script_hash (s, o.str());
 
       // See if we can use cached source/module.
-      if (get_from_cache(s))
+      if (get_script_from_cache(s))
         {
 	  // If our last pass isn't 5, we're done (since passes 3 and
 	  // 4 just generate what we just pulled out of the cache).
@@ -1352,6 +1352,12 @@ main (int argc, char * const argv [])
   times (& tms_before);
   gettimeofday (&tv_before, NULL);
   STAP_PROBE1(stap, pass4__start, &s);
+
+  if (s.use_cache)
+    {
+      find_stapconf_hash(s);
+      get_stapconf_from_cache(s);
+    }
   rc = compile_pass (s);
 
   if (rc == 0 && s.last_pass == 4)
@@ -1377,7 +1383,9 @@ main (int argc, char * const argv [])
     {
       // Update cache. Cache cleaning is kicked off at the beginning of this function.
       if (s.use_script_cache)
-        add_to_cache(s);
+        add_script_to_cache(s);
+      if (s.use_cache)
+        add_stapconf_to_cache(s);
 
       // We may need to save the module in $CWD if the cache was
       // inaccessible for some reason.
diff --git a/session.h b/session.h
index f45f6c77..ccf6c1da 100644
--- a/session.h
+++ b/session.h
@@ -22,6 +22,7 @@ extern "C" {
 
 
 // forward decls for all referenced systemtap types
+struct hash;
 struct match_node;
 struct stapfile;
 struct vardecl;
@@ -131,6 +132,7 @@ struct systemtap_session
   std::string cache_path;
   std::string hash_path;
   std::string stapconf_path;
+  hash *base_hash;
 
   // dwarfless operation
   bool consult_symtab;