This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCH 3/5] IPA ICF pass


> diff --git a/gcc/cgraph.h b/gcc/cgraph.h
> index fb41b01..2de98b4 100644
> --- a/gcc/cgraph.h
> +++ b/gcc/cgraph.h
> @@ -172,6 +172,12 @@ public:
>    /* Dump referring in list to FILE.  */
>    void dump_referring (FILE *);
>  
> +  /* Get number of references for this node.  */
> +  inline unsigned get_references_count (void)
> +  {
> +    return ref_list.references ? ref_list.references->length () : 0;
> +  }

Probably better called num_references() (like we have num_edge in basic-block.h)
> @@ -8068,6 +8069,19 @@ it may significantly increase code size
>  (see @option{--param ipcp-unit-growth=@var{value}}).
>  This flag is enabled by default at @option{-O3}.
>  
> +@item -fipa-icf
> +@opindex fipa-icf
> +Perform Identical Code Folding for functions and read-only variables.
> +The optimization reduces code size and may disturb unwind stacks by replacing
> +a function by equivalent one with a different name. The optimization works
> +more effectively with link time optimization enabled.
> +
> +Nevertheless the behavior is similar to Gold Linker ICF optimization, GCC ICF
> +works on different levels and thus the optimizations are not same - there are
> +equivalences that are found only by GCC and equivalences found only by Gold.
> +
> +This flag is enabled by default at @option{-O2}.
... and -Os?
> +    case ARRAY_REF:
> +    case ARRAY_RANGE_REF:
> +      {
> +	x1 = TREE_OPERAND (t1, 0);
> +	x2 = TREE_OPERAND (t2, 0);
> +	y1 = TREE_OPERAND (t1, 1);
> +	y2 = TREE_OPERAND (t2, 1);
> +
> +	if (!compare_operand (array_ref_low_bound (t1),
> +			      array_ref_low_bound (t2)))
> +	  return return_false_with_msg ("");
> +	if (!compare_operand (array_ref_element_size (t1),
> +			      array_ref_element_size (t2)))
> +	  return return_false_with_msg ("");
> +	if (!compare_operand (x1, x2))
> +	  return return_false_with_msg ("");
> +	return compare_operand (y1, y2);
> +      }

No need for {...} if there are no local vars.
> +bool
> +func_checker::compare_function_decl (tree t1, tree t2)
> +{
> +  bool ret = false;
> +
> +  if (t1 == t2)
> +    return true;
> +
> +  symtab_node *n1 = symtab_node::get (t1);
> +  symtab_node *n2 = symtab_node::get (t2);
> +
> +  if (m_ignored_source_nodes != NULL && m_ignored_target_nodes != NULL)
> +    {
> +      ret = m_ignored_source_nodes->contains (n1)
> +	    && m_ignored_target_nodes->contains (n2);
> +
> +      if (ret)
> +	return true;
> +    }
> +
> +  /* If function decl is WEAKREF, we compare targets.  */
> +  cgraph_node *f1 = cgraph_node::get (t1);
> +  cgraph_node *f2 = cgraph_node::get (t2);
> +
> +  if(f1 && f2 && f1->weakref && f2->weakref)
> +    ret = f1->alias_target == f2->alias_target;
> +
> +  return ret;

Comparing aliases is bit more complicated than just handling weakrefs. I have
patch for symtab_node::equivalent_address_p somewhre in queue.  lets just drop
the fancy stuff for the moment and compare f1&&f2 for equivalence.
> +  ret = compare_decl (t1, t2);

Why functions are not compared with compare_decl while variables are?
> +
> +  return return_with_debug (ret);
> +}
> +
> +void
> +func_checker::parse_labels (sem_bb *bb)
> +{
> +  for (gimple_stmt_iterator gsi = gsi_start_bb (bb->bb); !gsi_end_p (gsi);
> +       gsi_next (&gsi))
> +    {
> +      gimple stmt = gsi_stmt (gsi);
> +
> +      if (gimple_code (stmt) == GIMPLE_LABEL)
> +	{
> +	  tree t = gimple_label_label (stmt);
> +	  gcc_assert (TREE_CODE (t) == LABEL_DECL);
> +
> +	  m_label_bb_map.put (t, bb->bb->index);
> +	}
> +    }
> +}
> +
> +/* Basic block equivalence comparison function that returns true if
> +   basic blocks BB1 and BB2 (from functions FUNC1 and FUNC2) correspond.
> +
> +   In general, a collection of equivalence dictionaries is built for types
> +   like SSA names, declarations (VAR_DECL, PARM_DECL, ..). This infrastructure
> +   is utilized by every statement-by-stament comparison function.  */
> +
> +bool
> +func_checker::compare_bb (sem_bb *bb1, sem_bb *bb2)
> +{
> +  unsigned i;
> +  gimple_stmt_iterator gsi1, gsi2;
> +  gimple s1, s2;
> +
> +  if (bb1->nondbg_stmt_count != bb2->nondbg_stmt_count
> +      || bb1->edge_count != bb2->edge_count)
> +    return return_false ();
> +
> +  gsi1 = gsi_start_bb (bb1->bb);
> +  gsi2 = gsi_start_bb (bb2->bb);
> +
> +  for (i = 0; i < bb1->nondbg_stmt_count; i++)
> +    {
> +      if (is_gimple_debug (gsi_stmt (gsi1)))
> +	gsi_next_nondebug (&gsi1);
> +
> +      if (is_gimple_debug (gsi_stmt (gsi2)))
> +	gsi_next_nondebug (&gsi2);
> +
> +      s1 = gsi_stmt (gsi1);
> +      s2 = gsi_stmt (gsi2);
> +
> +      int eh1 = lookup_stmt_eh_lp_fn
> +		(DECL_STRUCT_FUNCTION (m_source_func_decl), s1);
> +      int eh2 = lookup_stmt_eh_lp_fn
> +		(DECL_STRUCT_FUNCTION (m_target_func_decl), s2);
> +
> +      if (eh1 != eh2)
> +	return return_false_with_msg ("EH regions are different");
> +
> +      if (gimple_code (s1) != gimple_code (s2))
> +	return return_false_with_msg ("gimple codes are different");
> +
> +      switch (gimple_code (s1))
> +	{
> +	case GIMPLE_CALL:
> +	  if (!compare_gimple_call (s1, s2))
> +	    return return_different_stmts (s1, s2, "GIMPLE_CALL");
> +	  break;
> +	case GIMPLE_ASSIGN:
> +	  if (!compare_gimple_assign (s1, s2))
> +	    return return_different_stmts (s1, s2, "GIMPLE_ASSIGN");
> +	  break;
> +	case GIMPLE_COND:
> +	  if (!compare_gimple_cond (s1, s2))
> +	    return return_different_stmts (s1, s2, "GIMPLE_COND");
> +	  break;
> +	case GIMPLE_SWITCH:
> +	  if (!compare_gimple_switch (s1, s2))
> +	    return return_different_stmts (s1, s2, "GIMPLE_SWITCH");
> +	  break;
> +	case GIMPLE_DEBUG:
> +	case GIMPLE_EH_DISPATCH:
> +	  break;
> +	case GIMPLE_RESX:
> +	  if (!compare_gimple_resx (s1, s2))
> +	    return return_different_stmts (s1, s2, "GIMPLE_RESX");
> +	  break;
> +	case GIMPLE_LABEL:
> +	  if (!compare_gimple_label (s1, s2))
> +	    return return_different_stmts (s1, s2, "GIMPLE_LABEL");
> +	  break;
> +	case GIMPLE_RETURN:
> +	  if (!compare_gimple_return (s1, s2))
> +	    return return_different_stmts (s1, s2, "GIMPLE_RETURN");
> +	  break;
> +	case GIMPLE_GOTO:
> +	  if (!compare_gimple_goto (s1, s2))
> +	    return return_different_stmts (s1, s2, "GIMPLE_GOTO");
> +	  break;
> +	case GIMPLE_ASM:
> +	  if (!compare_gimple_asm (s1, s2))
> +	    return return_different_stmts (s1, s2, "GIMPLE_ASM");
> +	  break;
> +	case GIMPLE_PREDICT:
> +	case GIMPLE_NOP:
> +	  return true;
> +	default:
> +	  return return_false_with_msg ("Unknown GIMPLE code reached");
> +	}
> +
> +      gsi_next (&gsi1);
> +      gsi_next (&gsi2);
> +    }
> +
> +  return true;
> +}
> +
> +/* Verifies for given GIMPLEs S1 and S2 that
> +   call statements are semantically equivalent.  */
> +
> +bool
> +func_checker::compare_gimple_call (gimple s1, gimple s2)
> +{
> +  unsigned i;
> +  tree t1, t2;
> +
> +  if (gimple_call_num_args (s1) != gimple_call_num_args (s2))
> +    return false;
> +
> +  t1 = gimple_call_fndecl (s1);
> +  t2 = gimple_call_fndecl (s2);
> +
> +  /* Function pointer variables are not supported yet.  */
> +  if (t1 == NULL || t2 == NULL)
> +    {
> +      if (!compare_operand (t1, t2))
> +	return return_false();

I think the comment above is out of date. compare_operand should do the right
job for indirect calls.
> +
> +  if (cn1 && cn2 && cn1->weakref && cn2->weakref
> +      && cn1->alias_target == cn2->alias_target)
> +    return true;

Lets consistently drop the weakrefs handling and add full alias handling incrementally.
> +
> +  /* Checking function arguments.  */
attributes
> +  tree decl1 = DECL_ATTRIBUTES (decl);
> +  tree decl2 = DECL_ATTRIBUTES (m_compared_func->decl);

You can still do this as part of the wap_comparison, right?
> +
> +  m_checker = new func_checker (decl, m_compared_func->decl,
> +				compare_polymorphic_p (),
> +				false,
> +				&refs_set,
> +				&m_compared_func->refs_set);
> +  while (decl1)
> +    {
> +      if (decl2 == NULL)
> +	return return_false ();
> +
> +      if (get_attribute_name (decl1) != get_attribute_name (decl2))
> +	return return_false ();
> +
> +      tree attr_value1 = TREE_VALUE (decl1);
> +      tree attr_value2 = TREE_VALUE (decl2);
> +
> +      if (attr_value1 && attr_value2)
> +	{
> +	  bool ret = m_checker->compare_operand (TREE_VALUE (attr_value1),
> +						 TREE_VALUE (attr_value2));
> +	  if (!ret)
> +	    return return_false_with_msg ("attribute values are different");
> +	}
> +      else if (!attr_value1 && !attr_value2)
> +	{}
> +      else
> +	return return_false ();
> +
> +      decl1 = TREE_CHAIN (decl1);
> +      decl2 = TREE_CHAIN (decl2);
> +    }
> +
> +  if (decl1 != decl2)
> +    return return_false();
> +
> +
> +  for (arg1 = DECL_ARGUMENTS (decl),
> +       arg2 = DECL_ARGUMENTS (m_compared_func->decl);
> +       arg1; arg1 = DECL_CHAIN (arg1), arg2 = DECL_CHAIN (arg2))
> +    if (!m_checker->compare_decl (arg1, arg2))
> +      return return_false ();
> +
> +  /* Fill-up label dictionary.  */
> +  for (unsigned i = 0; i < bb_sorted.length (); ++i)
> +    {
> +      m_checker->parse_labels (bb_sorted[i]);
> +      m_checker->parse_labels (m_compared_func->bb_sorted[i]);
> +    }
> +
> +  /* Checking all basic blocks.  */
> +  for (unsigned i = 0; i < bb_sorted.length (); ++i)
> +    if(!m_checker->compare_bb (bb_sorted[i], m_compared_func->bb_sorted[i]))
> +      return return_false();
> +
> +  dump_message ("All BBs are equal\n");
> +
> +  /* Basic block edges check.  */
> +  for (unsigned i = 0; i < bb_sorted.length (); ++i)
> +    {
> +      bb_dict = XNEWVEC (int, bb_sorted.length () + 2);
> +      memset (bb_dict, -1, (bb_sorted.length () + 2) * sizeof (int));
> +
> +      bb1 = bb_sorted[i]->bb;
> +      bb2 = m_compared_func->bb_sorted[i]->bb;
> +
> +      ei2 = ei_start (bb2->preds);
> +
> +      for (ei1 = ei_start (bb1->preds); ei_cond (ei1, &e1); ei_next (&ei1))
> +	{
> +	  ei_cond (ei2, &e2);
> +
> +	  if (e1->flags != e2->flags)
> +	    return return_false_with_msg ("flags comparison returns false");
> +
> +	  if (!bb_dict_test (bb_dict, e1->src->index, e2->src->index))
> +	    return return_false_with_msg ("edge comparison returns false");
> +
> +	  if (!bb_dict_test (bb_dict, e1->dest->index, e2->dest->index))
> +	    return return_false_with_msg ("BB comparison returns false");
> +
> +	  if (!m_checker->compare_edge (e1, e2))
> +	    return return_false_with_msg ("edge comparison returns false");
> +
> +	  ei_next (&ei2);
> +	}
> +    }
> +
> +  /* Basic block PHI nodes comparison.  */
> +  for (unsigned i = 0; i < bb_sorted.length (); i++)
> +    if (!compare_phi_node (bb_sorted[i]->bb, m_compared_func->bb_sorted[i]->bb))
> +      return return_false_with_msg ("PHI node comparison returns false");
> +
> +  return result;
> +}

The rest of patch seems fine.  I think we went across enough of iteraitons, the patch is OK
with changes above and lets handle rest incrementally.

Thanks!
Honza


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]