[gcc/devel/gccgo] tree-optimization/95045 - fix SM with exit exiting multiple loops

[Ian Lance Taylor]

https://gcc.gnu.org/g:bb63ca63e744c08bc5a9ffa53df62ea35f098b0b

commit bb63ca63e744c08bc5a9ffa53df62ea35f098b0b
Author: Richard Biener <rguenther@suse.de>
Date:   Mon May 11 15:26:09 2020 +0200

    tree-optimization/95045 - fix SM with exit exiting multiple loops
    
    Since we apply SM to an edge which exits multiple loops we have
    to make sure to commit insertions on it immediately since otherwise
    store order is not preserved.
    
    2020-05-12  Richard Biener  <rguenther@suse.de>
    
            PR tree-optimization/95045
            * dbgcnt.def (lim): Add debug-counter.
            * tree-ssa-loop-im.c: Include dbgcnt.h.
            (find_refs_for_sm): Use lim debug counter for store motion
            candidates.
            (do_store_motion): Rename form store_motion.  Commit edge
            insertions...
            (store_motion_loop): ... here.
            (tree_ssa_lim): Adjust.

Diff:
---
 gcc/ChangeLog          | 12 ++++++++++++
 gcc/dbgcnt.def         |  1 +
 gcc/tree-ssa-loop-im.c | 15 ++++++++++-----
 3 files changed, 23 insertions(+), 5 deletions(-)

diff --git a/gcc/ChangeLog b/gcc/ChangeLog
index ef6867527a1..9b8b5aee9f6 100644
--- a/gcc/ChangeLog
+++ b/gcc/ChangeLog
@@ -1,3 +1,15 @@
+2020-05-12  Richard Biener  <rguenther@suse.de>
+
+	PR tree-optimization/95045
+	* dbgcnt.def (lim): Add debug-counter.
+	* tree-ssa-loop-im.c: Include dbgcnt.h.
+	(find_refs_for_sm): Use lim debug counter for store motion
+	candidates.
+	(do_store_motion): Rename form store_motion.  Commit edge
+	insertions...
+	(store_motion_loop): ... here.
+	(tree_ssa_lim): Adjust.
+
 2020-05-11  Kelvin Nilsen  <kelvin@gcc.gnu.org>
 
 	* config/rs6000/altivec.h (vec_clzm): Rename to vec_cntlzm.
diff --git a/gcc/dbgcnt.def b/gcc/dbgcnt.def
index 232b1928983..3998c9636aa 100644
--- a/gcc/dbgcnt.def
+++ b/gcc/dbgcnt.def
@@ -174,6 +174,7 @@ DEBUG_COUNTER (ipa_sra_params)
 DEBUG_COUNTER (ipa_sra_retvalues)
 DEBUG_COUNTER (ira_move)
 DEBUG_COUNTER (ivopts_loop)
+DEBUG_COUNTER (lim)
 DEBUG_COUNTER (local_alloc_for_sched)
 DEBUG_COUNTER (match)
 DEBUG_COUNTER (merged_ipa_icf)
diff --git a/gcc/tree-ssa-loop-im.c b/gcc/tree-ssa-loop-im.c
index bb78dfb2ce8..0d77aaa08a5 100644
--- a/gcc/tree-ssa-loop-im.c
+++ b/gcc/tree-ssa-loop-im.c
@@ -47,6 +47,7 @@ along with GCC; see the file COPYING3.  If not see
 #include "alias.h"
 #include "builtins.h"
 #include "tree-dfa.h"
+#include "dbgcnt.h"
 
 /* TODO:  Support for predicated code motion.  I.e.
 
@@ -2862,7 +2863,7 @@ find_refs_for_sm (class loop *loop, bitmap sm_executed, bitmap refs_to_sm)
   EXECUTE_IF_AND_COMPL_IN_BITMAP (refs, sm_executed, 0, i, bi)
     {
       ref = memory_accesses.refs_list[i];
-      if (can_sm_ref_p (loop, ref))
+      if (can_sm_ref_p (loop, ref) && dbg_cnt (lim))
 	bitmap_set_bit (refs_to_sm, i);
     }
 }
@@ -2900,7 +2901,12 @@ store_motion_loop (class loop *loop, bitmap sm_executed)
     {
       find_refs_for_sm (loop, sm_executed, sm_in_loop);
       if (!bitmap_empty_p (sm_in_loop))
-	hoist_memory_references (loop, sm_in_loop, exits);
+	{
+	  hoist_memory_references (loop, sm_in_loop, exits);
+	  /* Commit edge inserts here to preserve the order of stores
+	     when an exit exits multiple loops.  */
+	  gsi_commit_edge_inserts ();
+	}
     }
   exits.release ();
 
@@ -2915,7 +2921,7 @@ store_motion_loop (class loop *loop, bitmap sm_executed)
    loops.  */
 
 static void
-store_motion (void)
+do_store_motion (void)
 {
   class loop *loop;
   bitmap sm_executed = BITMAP_ALLOC (&lim_bitmap_obstack);
@@ -2924,7 +2930,6 @@ store_motion (void)
     store_motion_loop (loop, sm_executed);
 
   BITMAP_FREE (sm_executed);
-  gsi_commit_edge_inserts ();
 }
 
 /* Fills ALWAYS_EXECUTED_IN information for basic blocks of LOOP, i.e.
@@ -3141,7 +3146,7 @@ tree_ssa_lim (void)
 
   /* Execute store motion.  Force the necessary invariants to be moved
      out of the loops as well.  */
-  store_motion ();
+  do_store_motion ();
 
   /* Move the expressions that are expensive enough.  */
   todo = move_computations ();