[PATCH 1/2, PR 61654] Handle newly truly expanded artificial_thunks

[PATCH 1/2, PR 61654] Handle newly truly expanded artificial_thunks

[Martin Jambor]

Hi,

I did not think it was possible, but it can happen that when
duplicate_thunk_for_node creates a duplicate of a thunk which
previously expand_thunk left alone to be expanded into assembly by the
back end, the newly created thunk does get expanded by expand_thunk.
When this happens, we end up with an un-analyzed node which triggers
an assert later on.

This patch deals with the situation by analyzing the newly expanded
thunk.  This revealed that DECL_ARGUMENTS were insufficiently copied
for the new decl and it was sharing them with the old one.  So this
patch fixes this as well.

Bootstrapped and tested on x86_64-linux and i686-linux (where the bug
triggered), OK for trunk and the 4.9 branch?

Thanks,

Martin


2014-09-01  Martin Jambor  <mjambor@suse.cz>

	PR ipa/61654
	* cgraphclones.c (duplicate_thunk_for_node): Copy arguments of the
	new decl properly.  Analyze the new thunk if it is expanded.

gcc/testsuite/
	* g++.dg/ipa/pr61654.C: New test.
---
 gcc/cgraphclones.c                 | 21 ++++++++++++++++++++
 gcc/testsuite/g++.dg/ipa/pr61654.C | 40 ++++++++++++++++++++++++++++++++++++++
 2 files changed, 61 insertions(+)
 create mode 100644 gcc/testsuite/g++.dg/ipa/pr61654.C

diff --git a/gcc/cgraphclones.c b/gcc/cgraphclones.c
index eb04418..2a17de5 100644
--- a/gcc/cgraphclones.c
+++ b/gcc/cgraphclones.c
@@ -334,6 +334,22 @@ duplicate_thunk_for_node (cgraph_node *thunk, cgraph_node *node)
 						node->clone.args_to_skip,
 						false);
     }
+
+  tree *link = &DECL_ARGUMENTS (new_decl);
+  int i = 0;
+  for (tree pd = DECL_ARGUMENTS (thunk->decl); pd; pd = DECL_CHAIN (pd), i++)
+    {
+      if (!node->clone.args_to_skip
+	  || !bitmap_bit_p (node->clone.args_to_skip, i))
+	{
+	  tree nd = copy_node (pd);
+	  DECL_CONTEXT (nd) = new_decl;
+	  *link = nd;
+	  link = &DECL_CHAIN (nd);
+	}
+    }
+  *link = NULL_TREE;
+
   gcc_checking_assert (!DECL_STRUCT_FUNCTION (new_decl));
   gcc_checking_assert (!DECL_INITIAL (new_decl));
   gcc_checking_assert (!DECL_RESULT (new_decl));
@@ -357,6 +373,11 @@ duplicate_thunk_for_node (cgraph_node *thunk, cgraph_node *node)
   symtab->call_edge_duplication_hooks (thunk->callees, e);
   if (!new_thunk->expand_thunk (false, false))
     new_thunk->analyzed = true;
+  else
+    {
+      new_thunk->thunk.thunk_p = false;
+      new_thunk->analyze ();
+    }
 
   symtab->call_cgraph_duplication_hooks (thunk, new_thunk);
   return new_thunk;
diff --git a/gcc/testsuite/g++.dg/ipa/pr61654.C b/gcc/testsuite/g++.dg/ipa/pr61654.C
new file mode 100644
index 0000000..d07e458
--- /dev/null
+++ b/gcc/testsuite/g++.dg/ipa/pr61654.C
@@ -0,0 +1,40 @@
+/* { dg-do compile } */
+/* { dg-options "-O3" } */
+
+/* The bug only presented itself on a 32 bit i386 but in theory it might also
+   pop up elsewhere and we do not want to put -m32 options to testcase
+   options.  */
+
+struct A
+{
+  virtual int a (int, int = 0) = 0;
+  void b ();
+  void c ();
+  int d;
+};
+
+struct B : virtual A
+{
+  int a (int, int);
+  int e;
+};
+
+int f;
+
+void
+A::b ()
+{
+  a (0);
+}
+
+void
+A::c ()
+{
+  a (f);
+}
+
+int
+B::a (int, int)
+{
+  return e;
+}
-- 
1.8.4.5

Re: [PATCH 1/2, PR 61654] Handle newly truly expanded artificial_thunks

[Jeff Law]

On 09/03/14 02:45, Martin Jambor wrote:
> Hi,
>
> I did not think it was possible, but it can happen that when
> duplicate_thunk_for_node creates a duplicate of a thunk which
> previously expand_thunk left alone to be expanded into assembly by the
> back end, the newly created thunk does get expanded by expand_thunk.
> When this happens, we end up with an un-analyzed node which triggers
> an assert later on.
>
> This patch deals with the situation by analyzing the newly expanded
> thunk.  This revealed that DECL_ARGUMENTS were insufficiently copied
> for the new decl and it was sharing them with the old one.  So this
> patch fixes this as well.
>
> Bootstrapped and tested on x86_64-linux and i686-linux (where the bug
> triggered), OK for trunk and the 4.9 branch?
>
> Thanks,
>
> Martin
>
>
> 2014-09-01  Martin Jambor  <mjambor@suse.cz>
>
> 	PR ipa/61654
> 	* cgraphclones.c (duplicate_thunk_for_node): Copy arguments of the
> 	new decl properly.  Analyze the new thunk if it is expanded.
>
> gcc/testsuite/
> 	* g++.dg/ipa/pr61654.C: New test.
OK.
Jeff

Re: [PATCH, 4.9, PR 61654] Handle newly truly expanded artificial_thunks

[Martin Jambor]

On Wed, Sep 03, 2014 at 10:45:34AM +0200, Martin Jambor wrote:
> Hi,
> 
> I did not think it was possible, but it can happen that when
> duplicate_thunk_for_node creates a duplicate of a thunk which
> previously expand_thunk left alone to be expanded into assembly by the
> back end, the newly created thunk does get expanded by expand_thunk.
> When this happens, we end up with an un-analyzed node which triggers
> an assert later on.
> 
> This patch deals with the situation by analyzing the newly expanded
> thunk.  This revealed that DECL_ARGUMENTS were insufficiently copied
> for the new decl and it was sharing them with the old one.  So this
> patch fixes this as well.
> 
> Bootstrapped and tested on x86_64-linux and i686-linux (where the bug
> triggered), OK for trunk and the 4.9 branch?
> 

After Jeff's approval, I have committed the patch to the trunk.  The
patch for 4.9 is however slightly larger, because before cgraph API
C++ification analyze function was a static function (now it is a
method of cgraph_node).  Because my fix needs to call it from another
compilation unit, I need to make it externally visible.  The 4.9 patch
is below, it also passes bootstrap and test on an x86_64-linux, i386
testing is underway.  Given that the original patch has been approved,
I intend to commit this one to the branch tomorrow if the test
finishes fine.

Thanks,

Martin


2014-09-10  Martin Jambor  <mjambor@suse.cz>

	PR ipa/61654
	* cgraph.h (cgraph_analyze_function): Declare.
	* cgraphunit.c: (analyze_function): Remove forward declaration,
	rename to cgraph_analyze_function, made external.
	* cgraphclones.c (duplicate_thunk_for_node): Copy arguments of the
        new decl properly.  Analyze the new thunk if it is expanded.

gcc/testsuite/
        * g++.dg/ipa/pr61654.C: New test.

diff --git a/gcc/cgraph.h b/gcc/cgraph.h
index 6c3be6d..0d13ebe 100644
--- a/gcc/cgraph.h
+++ b/gcc/cgraph.h
@@ -797,6 +797,7 @@ void cgraph_unnest_node (struct cgraph_node *);
 
 enum availability cgraph_function_body_availability (struct cgraph_node *);
 void cgraph_add_new_function (tree, bool);
+void cgraph_analyze_function (struct cgraph_node *);
 const char* cgraph_inline_failed_string (cgraph_inline_failed_t);
 cgraph_inline_failed_type_t cgraph_inline_failed_type (cgraph_inline_failed_t);
 
diff --git a/gcc/cgraphclones.c b/gcc/cgraphclones.c
index 972ca07..9ad76dd 100644
--- a/gcc/cgraphclones.c
+++ b/gcc/cgraphclones.c
@@ -334,6 +334,22 @@ duplicate_thunk_for_node (cgraph_node *thunk, cgraph_node *node)
 						node->clone.args_to_skip,
 						false);
     }
+
+  tree *link = &DECL_ARGUMENTS (new_decl);
+  int i = 0;
+  for (tree pd = DECL_ARGUMENTS (thunk->decl); pd; pd = DECL_CHAIN (pd), i++)
+    {
+      if (!node->clone.args_to_skip
+	  || !bitmap_bit_p (node->clone.args_to_skip, i))
+	{
+	  tree nd = copy_node (pd);
+	  DECL_CONTEXT (nd) = new_decl;
+	  *link = nd;
+	  link = &DECL_CHAIN (nd);
+	}
+    }
+  *link = NULL_TREE;
+
   gcc_checking_assert (!DECL_STRUCT_FUNCTION (new_decl));
   gcc_checking_assert (!DECL_INITIAL (new_decl));
   gcc_checking_assert (!DECL_RESULT (new_decl));
@@ -358,6 +374,11 @@ duplicate_thunk_for_node (cgraph_node *thunk, cgraph_node *node)
   cgraph_call_edge_duplication_hooks (thunk->callees, e);
   if (!expand_thunk (new_thunk, false))
     new_thunk->analyzed = true;
+  else
+    {
+      new_thunk->thunk.thunk_p = false;
+      cgraph_analyze_function (new_thunk);
+    }
   cgraph_call_node_duplication_hooks (thunk, new_thunk);
   return new_thunk;
 }
diff --git a/gcc/cgraphunit.c b/gcc/cgraphunit.c
index 8abdc5d..f486055 100644
--- a/gcc/cgraphunit.c
+++ b/gcc/cgraphunit.c
@@ -219,7 +219,6 @@ cgraph_node_set cgraph_new_nodes;
 static void expand_all_functions (void);
 static void mark_functions_to_output (void);
 static void expand_function (struct cgraph_node *);
-static void analyze_function (struct cgraph_node *);
 static void handle_alias_pairs (void);
 
 FILE *cgraph_dump_file;
@@ -331,7 +330,7 @@ cgraph_process_new_functions (void)
 
 	  gimple_register_cfg_hooks ();
 	  if (!node->analyzed)
-	    analyze_function (node);
+	    cgraph_analyze_function (node);
 	  push_cfun (DECL_STRUCT_FUNCTION (fndecl));
 	  if (cgraph_state == CGRAPH_STATE_IPA_SSA
 	      && !gimple_in_ssa_p (DECL_STRUCT_FUNCTION (fndecl)))
@@ -541,7 +540,7 @@ cgraph_add_new_function (tree fndecl, bool lowered)
 	if (lowered)
 	  node->lowered = true;
 	node->definition = true;
-	analyze_function (node);
+	cgraph_analyze_function (node);
 	push_cfun (DECL_STRUCT_FUNCTION (fndecl));
 	gimple_register_cfg_hooks ();
 	bitmap_obstack_initialize (NULL);
@@ -598,8 +597,8 @@ output_asm_statements (void)
 }
 
 /* Analyze the function scheduled to be output.  */
-static void
-analyze_function (struct cgraph_node *node)
+void
+cgraph_analyze_function (struct cgraph_node *node)
 {
   tree decl = node->decl;
   location_t saved_loc = input_location;
@@ -1014,7 +1013,7 @@ analyze_functions (void)
 		}
 
 	      if (!cnode->analyzed)
-		analyze_function (cnode);
+		cgraph_analyze_function (cnode);
 
 	      for (edge = cnode->callees; edge; edge = edge->next_callee)
 		if (edge->callee->definition)
diff --git a/gcc/testsuite/g++.dg/ipa/pr61654.C b/gcc/testsuite/g++.dg/ipa/pr61654.C
new file mode 100644
index 0000000..d07e458
--- /dev/null
+++ b/gcc/testsuite/g++.dg/ipa/pr61654.C
@@ -0,0 +1,40 @@
+/* { dg-do compile } */
+/* { dg-options "-O3" } */
+
+/* The bug only presented itself on a 32 bit i386 but in theory it might also
+   pop up elsewhere and we do not want to put -m32 options to testcase
+   options.  */
+
+struct A
+{
+  virtual int a (int, int = 0) = 0;
+  void b ();
+  void c ();
+  int d;
+};
+
+struct B : virtual A
+{
+  int a (int, int);
+  int e;
+};
+
+int f;
+
+void
+A::b ()
+{
+  a (0);
+}
+
+void
+A::c ()
+{
+  a (f);
+}
+
+int
+B::a (int, int)
+{
+  return e;
+}