| Submitter | Jason Merrill |
|---|---|
| Date | Dec. 3, 2010, 4:19 p.m. |
| Message ID | <4CF91888.4020301@redhat.com> |
| Download | mbox | patch |
| Permalink | /patch/74161/ |
| State | New |
| Headers | show |
Comments
Patch
diff --git a/gcc/cp/semantics.c b/gcc/cp/semantics.c index 9b565da..2424e49 100644 --- a/gcc/cp/semantics.c +++ b/gcc/cp/semantics.c @@ -5484,10 +5484,8 @@ build_data_member_initialization (tree t, VEC(constructor_elt,gc) **vec) init = unshare_expr (TREE_OPERAND (t, 1)); if (TREE_CODE (member) == INDIRECT_REF) { - /* Don't put out anything for value-init of an empty base. */ + /* Don't put out anything for an empty base. */ gcc_assert (is_empty_class (TREE_TYPE (member))); - gcc_assert (TREE_CODE (init) == CONSTRUCTOR - && CONSTRUCTOR_NELTS (init) == 0); return true; } }
The asserts in build_data_member_initialization were too strict; we can see an initialization of an empty base in a copy constructor, too. Tested x86_64-pc-linux-gnu, applied to trunk. commit 70d4870d0fd8c4bc580915f16f1e94641e346a3f Author: Jason Merrill <jason@redhat.com> Date: Wed Dec 1 17:43:49 2010 -0500 * semantics.c (build_data_member_initialization): Remove assert.