[Bug c++/101480] [11/12 Regression] Miscompiled code involving operator new

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101480

--- Comment #14 from Richard Biener <rguenth at gcc dot gnu.org> ---
diff --git a/gcc/gimple.c b/gcc/gimple.c
index 863bc0d17f1..e085d9de49a 100644
--- a/gcc/gimple.c
+++ b/gcc/gimple.c
@@ -1516,12 +1516,12 @@ gimple_call_fnspec (const gcall *stmt)
       && DECL_IS_OPERATOR_DELETE_P (fndecl)
       && DECL_IS_REPLACEABLE_OPERATOR (fndecl)
       && gimple_call_from_new_or_delete (stmt))
-    return ".co ";
+    return ". o ";
   /* Similarly operator new can be treated as malloc.  */
   if (fndecl
       && DECL_IS_REPLACEABLE_OPERATOR_NEW_P (fndecl)
       && gimple_call_from_new_or_delete (stmt))
-    return "mC";
+    return "m ";
   return "";
 }


regresses

FAIL: g++.dg/warn/Warray-bounds-16.C  -std=gnu++14  scan-tree-dump-not
optimized "goto"
FAIL: g++.dg/warn/Warray-bounds-16.C  -std=gnu++14 (test for excess errors)

where we now diagnose

/home/rguenther/src/trunk/gcc/testsuite/g++.dg/warn/Warray-bounds-16.C:22:7:
warning: array subscript 0 is outside array bounds of 'void [0]'
[-Warray-bounds]
/home/rguenther/src/trunk/gcc/testsuite/g++.dg/warn/Warray-bounds-16.C:22:7:
warning: 'void* __builtin_memset(void*, int, long unsigned int)' offset [0, 3]
is out of the bounds [0, 0] [-Warray-bounds]

the testcase does

    m = i;
    p = (int*) new unsigned char [sizeof (int) * m];

    for (int i = 0; i < m; i++)
      new (p + i) int ();

and we likely have to assume that 'new' changes 'm'.