public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
* [PATCH, RFC] Add a pass counter for "are we there yet" purposes
@ 2017-10-16  5:46 Sandra Loosemore
  2017-10-16  8:38 ` Richard Biener
  0 siblings, 1 reply; 8+ messages in thread
From: Sandra Loosemore @ 2017-10-16  5:46 UTC (permalink / raw)
  To: gcc-patches; +Cc: mjambor, Jeff Law, Jakub Jelinek

[-- Attachment #1: Type: text/plain, Size: 2041 bytes --]

This patch is a first cut at solving the problem discussed in this thread

https://gcc.gnu.org/ml/gcc/2017-10/msg00016.html

where I have some nios2 backend patches in my queue that need a way of 
knowing whether the split1 pass has run yet.  There seemed to be 
agreement that a general way to query the pass manager for this 
information would be useful.

The approach I took here is to add a new counter field, so that I can do 
the test I want with

   opt_pass *split1_pass = g->get_passes ()->get_pass_split_all_insns ();
   if (current_pass->pass_number > split1_pass->pass_number)
     ...

Well, mostly.  :-P  There are some gotchas.

* TARGET_ASM_OUTPUT_MI_THUNK is called outside the pass manager (so 
current_pass is NULL) and on many targets this hook is implemented by 
setting reload_completed to 1, generating some RTL, and invoking some 
passes directly to emit code.

* modulo-sched.c also plays tricks with setting reload_completed to 
pretend to be something it's not.

* Possibly other places?  E.g. I'm not familiar with how plugins work.

For my purposes it's good enough to check reload_completed before the 
test in the code snippet above, but trying to determine whether a 
particular post-reload pass has run won't work.  So this isn't as 
general as it ought to be, at least not until we get rid of the 
reload_completed hackery.

Since this patch isn't useful without something that uses the pass 
counters, I tested it on nios2-linux-gnu with my not-yet-posted patch 
set, by wiring it up in parallel with my previously-implemented solution 
of adding a target-specific pass to set a flag, with various assertions 
to check for consistency.  I also had some temporary debugging code in 
there at one point to print the pass numbers.

WDYT?  Is this the right direction?  I'm somewhat worried that we're 
getting late in stage 1 and I'd really like to finish up my nios2 
patches; so if getting this right looks like a tar pit, I think I should 
probably stick with my previous implementation for now.

-Sandra


[-- Attachment #2: enumpasses.log --]
[-- Type: text/x-log, Size: 349 bytes --]

2017-10-15  Sandra Loosemore  <sandra@codesourcery.com>

	gcc/
	* pass_manager.h (pass_manager::enumerate_passes): Declare.
	* tree-pass.h (opt_pass::pass_number): Declare.
	* passes.c (enumerate_pass_list): New.
	(pass_manager::enumerate_passes): New.
	(pass_manager::register_pass): Call enumerate_passes.
	(pass_manager::pass_manager): Likewise.

[-- Attachment #3: enumpasses.patch --]
[-- Type: text/x-patch, Size: 2647 bytes --]

Index: gcc/pass_manager.h
===================================================================
--- gcc/pass_manager.h	(revision 253327)
+++ gcc/pass_manager.h	(working copy)
@@ -103,6 +103,7 @@ private:
   void set_pass_for_id (int id, opt_pass *pass);
   void register_dump_files (opt_pass *pass);
   void create_pass_tab () const;
+  void enumerate_passes () const;
 
 private:
   context *m_ctxt;
Index: gcc/tree-pass.h
===================================================================
--- gcc/tree-pass.h	(revision 253327)
+++ gcc/tree-pass.h	(working copy)
@@ -108,6 +108,11 @@ public:
   /* Static pass number, used as a fragment of the dump file name.  */
   int static_pass_number;
 
+  /* Pass number used for enumerating the pass list.  This differs from
+     static_pass_number in that the numbering is sequential and all passes
+     have a meaningful number assigned.  */
+  int pass_number;
+
 protected:
   gcc::context *m_ctxt;
 };
Index: gcc/passes.c
===================================================================
--- gcc/passes.c	(revision 253327)
+++ gcc/passes.c	(working copy)
@@ -899,6 +899,35 @@ pass_manager::create_pass_tab (void) con
 
 static bool override_gate_status (opt_pass *, tree, bool);
 
+/* Enumerate pass list PASS starting with counter COUNTER.  */
+
+static int
+enumerate_pass_list (opt_pass *pass, int counter)
+{
+  do
+    {
+      pass->pass_number = ++counter;
+      if (pass->sub)
+	counter = enumerate_pass_list (pass->sub, counter);
+      pass = pass->next;
+    }
+  while (pass);
+  return counter;
+}
+
+/* Enumerate all optimization passes.  */
+
+void
+pass_manager::enumerate_passes () const
+{
+  int counter = 0;
+  counter = enumerate_pass_list (all_lowering_passes, counter);
+  counter = enumerate_pass_list (all_small_ipa_passes, counter);
+  counter = enumerate_pass_list (all_regular_ipa_passes, counter);
+  counter = enumerate_pass_list (all_late_ipa_passes, counter);
+  counter = enumerate_pass_list (all_passes, counter);
+}
+
 /* Dump the instantiated name for PASS. IS_ON indicates if PASS
    is turned on or not.  */
 
@@ -1508,6 +1537,9 @@ pass_manager::register_pass (struct regi
       XDELETE (added_pass_nodes);
       added_pass_nodes = next_node;
     }
+
+  /* Update the pass enumeration.  */
+  enumerate_passes ();
 }
 
 /* Construct the pass tree.  The sequencing of passes is driven by
@@ -1614,6 +1646,9 @@ pass_manager::pass_manager (context *ctx
   register_dump_files (all_regular_ipa_passes);
   register_dump_files (all_late_ipa_passes);
   register_dump_files (all_passes);
+
+  /* Enumerate passes.  */
+  enumerate_passes ();
 }
 
 static void

^ permalink raw reply	[flat|nested] 8+ messages in thread
end of thread, other threads:[~2017-10-23 11:00 UTC | newest]

Thread overview: 8+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2017-10-16  5:46 [PATCH, RFC] Add a pass counter for "are we there yet" purposes Sandra Loosemore
2017-10-16  8:38 ` Richard Biener
2017-10-16 16:00   ` Sandra Loosemore
2017-10-16 16:27     ` Richard Biener
2017-10-20  4:03       ` Sandra Loosemore
2017-10-23 10:21       ` Martin Jambor
2017-10-23 10:27         ` Jakub Jelinek
2017-10-23 11:07         ` Richard Biener

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).