On 8/6/19 2:42 PM, Martin Liška wrote:
> On 8/5/19 3:46 PM, Marc Glisse wrote:
>> On Mon, 5 Aug 2019, Martin Liška wrote:
>>
>>> You are right. It can really lead to confusion of the DCE.
>>>
>>> What we have is DECL_ABSTRACT_ORIGIN(decl) which we can use to indicate
>>> operators
>>> that were somehow modified by an IPA optimization.
>>
>> Looks similar to the cgraph_node->clone_of that Richard was talking about. I
>> have no idea which one would be best.
>
>
> Hm, strange that the ISRA clones don't have n->clone_of set. It's created
> here:
>
> #0 cgraph_node::create (decl=<function_decl 0x7ffff721c300
> _ZN1AdlEPvd.isra.0>) at /home/marxin/Programming/gcc/gcc/profile-count.h:751
> #1 0x0000000000bc8342 in cgraph_node::create_version_clone
> (this=<cgraph_node * const 0x7ffff7208000 "operator delete"/64>,
> new_decl=<optimized out>, redirect_callers=..., bbs_to_copy=0x0,
> suffix=0x1b74427 "isra") at
> /home/marxin/Programming/gcc/gcc/cgraphclones.c:960
> #2 0x0000000000bc9b9a in cgraph_node::create_version_clone_with_body
> (this=this@entry=<cgraph_node * const 0x7ffff7208000 "operator delete"/64>,
> redirect_callers=redirect_callers@entry=..., tree_map=tree_map@entry=0x0,
> args_to_skip=args_to_skip@entry=0x0,
> skip_return=skip_return@entry=false, bbs_to_copy=bbs_to_copy@entry=0x0,
> new_entry_block=<optimized out>, suffix=<optimized out>,
> target_attributes=<optimized out>) at
> /home/marxin/Programming/gcc/gcc/cgraphclones.c:1071
> #3 0x00000000010e4611 in modify_function (adjustments=..., node=<cgraph_node
> * 0x7ffff7208000 "operator delete"/64>) at
> /home/marxin/Programming/gcc/gcc/tree-sra.c:5493
> #4 ipa_early_sra () at /home/marxin/Programming/gcc/gcc/tree-sra.c:5731
> #5 (anonymous namespace)::pass_early_ipa_sra::execute (this=<optimized out>)
> at /home/marxin/Programming/gcc/gcc/tree-sra.c:5778
>
> @Martin, @Honza: Why do we not set clone_of in this transformation?
If I'm correct cgraph_node::clone is used for inline clones only?
Anyway, I'm sending patch that considers only such new/delete operators
that are not a clone of an original type. That should make the current
DCE code more solid.
Patch can bootstrap on x86_64-linux-gnu and survives regression tests.
Ready to be installed?
Thanks,
Martin
>
>>
>>> Do you believe it will be sufficient?
>>
>> In DCE only consider the operator delete that are not clones? (possibly the
>> same for operator new? I don't know how much we can change the return value
>> of a function in a clone) I guess that should work, and it wouldn't impact
>> the common case with default, global operator new/delete.
>
> I tent to limit that the only cgraph nodes that are not clones. I'm going to
> prepare a patch for it.
>
>>
>> An alternative would be to clear the DECL_IS_OPERATOR_DELETE flag when
>> creating a clone (possibly only if it modified the first parameter?). There
>> is probably not much information in the fact that a function that doesn't
>> even take a pointer used to be an operator delete.
>>
>>
>> By the way, I just thought of something, now that we handle class-specific
>> operator new/delete (but you could do the same with the global replacable
>> ones as well):
>>
>> #include <stdio.h>
>> int count = 0;
>> struct A {
>> __attribute__((malloc,noinline))
>> static void* operator new(unsigned long sz){++count;return ::operator
>> new(sz);}
>> static void operator delete(void* ptr){--count;::operator delete(ptr);}
>> };
>> int main(){
>> delete new A;
>> printf("%d\n",count);
>> }
>>
>> If we do not inline anything, we can remove the pair and nothing touches
>> count. If we inline both new and delete, we can then remove the inner pair
>> instead, count increases and decreases, fine. If we inline only one of them,
>> and DCE the mismatched pair new/delete, we get something inconsistent (count
>> is -1).
>>
>> This seems to indicate we should check that the new and delete match
>> somehow...
>>
>
>From 94110a47f0c13424d2e39d8f14bcc896a6097bd3 Mon Sep 17 00:00:00 2001
From: Martin Liska <[email protected]>
Date: Tue, 6 Aug 2019 16:14:48 +0200
Subject: [PATCH] Detect not-cloned new/delete operators in DCE.
gcc/ChangeLog:
2019-08-06 Martin Liska <[email protected]>
* gimple.c (gimple_call_operator_delete_p): Remove.
* gimple.h (gimple_call_operator_delete_p): Likewise.
* tree-ssa-dce.c (operator_new_candidate_p): New.
(operator_delete_candidate_p): Likewise.
(mark_stmt_if_obviously_necessary): Use operator_new_candidate_p
and operator_delete_candidate_p in order to detect operators
that are not not clones.
(mark_all_reaching_defs_necessary_1): Likewise.
(propagate_necessity): Likewise.
(eliminate_unnecessary_stmts): Likewise.
---
gcc/gimple.c | 12 ----------
gcc/gimple.h | 1 -
gcc/tree-ssa-dce.c | 59 ++++++++++++++++++++++++++--------------------
3 files changed, 34 insertions(+), 38 deletions(-)
diff --git a/gcc/gimple.c b/gcc/gimple.c
index 633ef512a19..684b8831b4d 100644
--- a/gcc/gimple.c
+++ b/gcc/gimple.c
@@ -2707,18 +2707,6 @@ gimple_builtin_call_types_compatible_p (const gimple *stmt, tree fndecl)
return true;
}
-/* Return true when STMT is operator delete call. */
-
-bool
-gimple_call_operator_delete_p (const gcall *stmt)
-{
- tree fndecl;
-
- if ((fndecl = gimple_call_fndecl (stmt)) != NULL_TREE)
- return DECL_IS_OPERATOR_DELETE_P (fndecl);
- return false;
-}
-
/* Return true when STMT is builtins call. */
bool
diff --git a/gcc/gimple.h b/gcc/gimple.h
index 55f5d0d33d9..7a1e1f49099 100644
--- a/gcc/gimple.h
+++ b/gcc/gimple.h
@@ -1548,7 +1548,6 @@ extern alias_set_type gimple_get_alias_set (tree);
extern bool gimple_ior_addresses_taken (bitmap, gimple *);
extern bool gimple_builtin_call_types_compatible_p (const gimple *, tree);
extern combined_fn gimple_call_combined_fn (const gimple *);
-extern bool gimple_call_operator_delete_p (const gcall *);
extern bool gimple_call_builtin_p (const gimple *);
extern bool gimple_call_builtin_p (const gimple *, enum built_in_class);
extern bool gimple_call_builtin_p (const gimple *, enum built_in_function);
diff --git a/gcc/tree-ssa-dce.c b/gcc/tree-ssa-dce.c
index afb7bd9dedc..05d30e6c839 100644
--- a/gcc/tree-ssa-dce.c
+++ b/gcc/tree-ssa-dce.c
@@ -114,6 +114,25 @@ static bool cfg_altered;
/* When non-NULL holds map from basic block index into the postorder. */
static int *bb_postorder;
+/* Return true when FNDECL is a new operator and not a clone. */
+
+static bool
+operator_new_candidate_p (tree fndecl)
+{
+ return (fndecl != NULL_TREE
+ && DECL_IS_OPERATOR_NEW_P (fndecl)
+ && DECL_ABSTRACT_ORIGIN (fndecl) == NULL_TREE);
+}
+
+/* Return true when FNDECL is a delete operator and not a clone. */
+
+static bool
+operator_delete_candidate_p (tree fndecl)
+{
+ return (fndecl != NULL_TREE
+ && DECL_IS_OPERATOR_DELETE_P (fndecl)
+ && DECL_ABSTRACT_ORIGIN (fndecl) == NULL_TREE);
+}
/* True if we should treat any stmt with a vdef as necessary. */
@@ -248,7 +267,7 @@ mark_stmt_if_obviously_necessary (gimple *stmt, bool aggressive)
if (callee != NULL_TREE
&& flag_allocation_dce
- && DECL_IS_REPLACEABLE_OPERATOR_NEW_P (callee))
+ && operator_new_candidate_p (callee))
return;
/* Most, but not all function calls are required. Function calls that
@@ -613,8 +632,8 @@ mark_all_reaching_defs_necessary_1 (ao_ref *ref ATTRIBUTE_UNUSED,
}
if (callee != NULL_TREE
- && (DECL_IS_REPLACEABLE_OPERATOR_NEW_P (callee)
- || DECL_IS_OPERATOR_DELETE_P (callee)))
+ && (operator_new_candidate_p (callee)
+ || operator_delete_candidate_p (callee)))
return false;
}
@@ -806,15 +825,10 @@ propagate_necessity (bool aggressive)
processing the argument. */
bool is_delete_operator
= (is_gimple_call (stmt)
- && gimple_call_operator_delete_p (as_a <gcall *> (stmt)));
+ && operator_delete_candidate_p (gimple_call_fndecl (as_a <gcall *> (stmt))));
if (is_delete_operator
|| gimple_call_builtin_p (stmt, BUILT_IN_FREE))
{
- /* It can happen that a user delete operator has the pointer
- argument optimized out already. */
- if (gimple_call_num_args (stmt) == 0)
- continue;
-
tree ptr = gimple_call_arg (stmt, 0);
gimple *def_stmt;
tree def_callee;
@@ -827,7 +841,7 @@ propagate_necessity (bool aggressive)
&& (DECL_FUNCTION_CODE (def_callee) == BUILT_IN_ALIGNED_ALLOC
|| DECL_FUNCTION_CODE (def_callee) == BUILT_IN_MALLOC
|| DECL_FUNCTION_CODE (def_callee) == BUILT_IN_CALLOC))
- || DECL_IS_REPLACEABLE_OPERATOR_NEW_P (def_callee)))
+ || operator_new_candidate_p (def_callee)))
{
/* Delete operators can have alignment and (or) size as next
arguments. When being a SSA_NAME, they must be marked
@@ -900,8 +914,8 @@ propagate_necessity (bool aggressive)
continue;
if (callee != NULL_TREE
- && (DECL_IS_REPLACEABLE_OPERATOR_NEW_P (callee)
- || DECL_IS_OPERATOR_DELETE_P (callee)))
+ && (operator_new_candidate_p (callee)
+ || operator_delete_candidate_p (callee)))
continue;
/* Calls implicitly load from memory, their arguments
@@ -1326,20 +1340,15 @@ eliminate_unnecessary_stmts (void)
if (gimple_plf (stmt, STMT_NECESSARY)
&& (gimple_call_builtin_p (stmt, BUILT_IN_FREE)
|| (is_gimple_call (stmt)
- && gimple_call_operator_delete_p (as_a <gcall *> (stmt)))))
+ && operator_delete_candidate_p (gimple_call_fndecl (as_a <gcall *> (stmt))))))
{
- /* It can happen that a user delete operator has the pointer
- argument optimized out already. */
- if (gimple_call_num_args (stmt) > 0)
+ tree ptr = gimple_call_arg (stmt, 0);
+ if (TREE_CODE (ptr) == SSA_NAME)
{
- tree ptr = gimple_call_arg (stmt, 0);
- if (TREE_CODE (ptr) == SSA_NAME)
- {
- gimple *def_stmt = SSA_NAME_DEF_STMT (ptr);
- if (!gimple_nop_p (def_stmt)
- && !gimple_plf (def_stmt, STMT_NECESSARY))
- gimple_set_plf (stmt, STMT_NECESSARY, false);
- }
+ gimple *def_stmt = SSA_NAME_DEF_STMT (ptr);
+ if (!gimple_nop_p (def_stmt)
+ && !gimple_plf (def_stmt, STMT_NECESSARY))
+ gimple_set_plf (stmt, STMT_NECESSARY, false);
}
}
@@ -1394,7 +1403,7 @@ eliminate_unnecessary_stmts (void)
&& DECL_FUNCTION_CODE (call) != BUILT_IN_CALLOC
&& !ALLOCA_FUNCTION_CODE_P
(DECL_FUNCTION_CODE (call))))
- && !DECL_IS_REPLACEABLE_OPERATOR_NEW_P (call))))
+ && !operator_new_candidate_p (call))))
{
something_changed = true;
if (dump_file && (dump_flags & TDF_DETAILS))
--
2.22.0
