On 1/17/22 03:56, Jakub Jelinek wrote:
Hi!

The following patch is miscompiled, cp_genericize_target_expr expects
that for the constant part split_nonconstant_init will emit an INIT_EXPR
that will initialize it, but that doesn't happen and instead we get
DECL_INITIAL on the TARGET_EXPR_SLOT that isn't initialized anywhere
in the IL.

The problem is that the TARGET_EXPR has been created while
current_function_decl was NULL, it is inside a global var initializer.
That means the build_local_temp created VAR_DECL has NULL DECL_CONTEXT.
Later on when genericizing the ssdf (current_function_decl is already
non-NULL), the new cp_genericize_target_expr is called and during
split_nonconstant_init it checks is_local_temp, but that due to the NULL
DECL_CONTEXT returns false.  DECL_CONTEXT is set only later on during
gimplification.

The following patch fixes it by setting DECL_CONTEXT also inside of
cp_genericize_target_expr, which fixes the testcase.  But if there are
better spots to do that, please let me know...

Bootstrapped/regtested on x86_64-linux and i686-linux, ok for trunk?

2022-01-17  Jakub Jelinek  <ja...@redhat.com>

        PR c++/104031
        * cp-gimplify.c (cp_genericize_target_expr): Set DECL_CONTEXT of
        TARGET_EXPR_SLOT to current_function_decl if it was NULL.

        * g++.dg/cpp1y/pr104031.C: New test.

--- gcc/cp/cp-gimplify.c.jj     2022-01-14 12:09:13.003250834 +0100
+++ gcc/cp/cp-gimplify.c        2022-01-16 21:34:52.300937996 +0100
@@ -930,6 +930,11 @@ cp_genericize_init_expr (tree *stmt_p)
  static void
  cp_genericize_target_expr (tree *stmt_p)
  {
+  /* If TARGET_EXPR is created for some global var initializer, the slot
+     will have NULL and so is_local_temp will return false for it.  If
+     this is a ssdf, set DECL_CONTEXT now.  */
+  if (DECL_CONTEXT (TARGET_EXPR_SLOT (*stmt_p)) == NULL_TREE)
+    DECL_CONTEXT (TARGET_EXPR_SLOT (*stmt_p)) = current_function_decl;
    cp_genericize_init (&TARGET_EXPR_INITIAL (*stmt_p),
                      TARGET_EXPR_INITIAL (*stmt_p),
                      TARGET_EXPR_SLOT (*stmt_p));

We could use a local variable for TARGET_EXPR_SLOT (*stmt_p) instead of repeating it. OK either way.

--- gcc/testsuite/g++.dg/cpp1y/pr104031.C.jj    2022-01-16 21:30:14.977838079 
+0100
+++ gcc/testsuite/g++.dg/cpp1y/pr104031.C       2022-01-16 21:30:46.028401294 
+0100
@@ -0,0 +1,23 @@
+// PR c++/104031
+// { dg-do run { target c++14 } }
+// { dg-options "-O2" }
+
+struct A {
+  A () {}
+  ~A () {}
+};
+struct B {
+  A a;
+  int b = 0;
+};
+struct C
+{
+  [[gnu::noipa]]
+  C (B x) { if (x.b != 42) __builtin_abort (); }
+};
+static C c ({ .a = A{}, .b = 42 });
+
+int
+main ()
+{
+}

        Jakub


Reply via email to