* [PATCH] c++: Treat unnamed bitfields as padding for __has_unique_object_representations [PR109096]
@ 2023-03-14 7:58 Jakub Jelinek
2023-03-14 14:35 ` Jason Merrill
0 siblings, 1 reply; 2+ messages in thread
From: Jakub Jelinek @ 2023-03-14 7:58 UTC (permalink / raw)
To: Jason Merrill; +Cc: gcc-patches
Hi!
As reported in the PR, for __has_unique_object_representations we
were treating unnamed bitfields as named ones, which is wrong, they
are actually padding.
THe following patch fixes that. Ok for trunk (and what about release
branches later?)?
2023-03-14 Jakub Jelinek <jakub@redhat.com>
PR c++/109096
* tree.cc (record_has_unique_obj_representations): Ignore unnamed
bitfields.
* g++.dg/cpp1z/has-unique-obj-representations3.C: New test.
--- gcc/cp/tree.cc.jj 2023-03-10 10:06:40.247560614 +0100
+++ gcc/cp/tree.cc 2023-03-13 10:38:03.394836926 +0100
@@ -4851,7 +4851,7 @@ record_has_unique_obj_representations (c
DECL_SIZE (field)))
return false;
}
- else if (DECL_C_BIT_FIELD (field))
+ else if (DECL_C_BIT_FIELD (field) && !DECL_UNNAMED_BIT_FIELD (field))
{
tree btype = DECL_BIT_FIELD_TYPE (field);
if (!type_has_unique_obj_representations (btype))
@@ -4862,7 +4862,7 @@ record_has_unique_obj_representations (c
offset_int cur = 0;
for (tree field = TYPE_FIELDS (t); field; field = DECL_CHAIN (field))
- if (TREE_CODE (field) == FIELD_DECL)
+ if (TREE_CODE (field) == FIELD_DECL && !DECL_UNNAMED_BIT_FIELD (field))
{
offset_int fld = wi::to_offset (DECL_FIELD_OFFSET (field));
offset_int bitpos = wi::to_offset (DECL_FIELD_BIT_OFFSET (field));
--- gcc/testsuite/g++.dg/cpp1z/has-unique-obj-representations3.C.jj 2023-03-13 10:50:21.705127719 +0100
+++ gcc/testsuite/g++.dg/cpp1z/has-unique-obj-representations3.C 2023-03-13 10:50:17.751185067 +0100
@@ -0,0 +1,10 @@
+// PR c++/109096
+// { dg-do compile { target c++11 } }
+
+#define INTB (__SIZEOF_INT__ * __CHAR_BIT__)
+struct U { int i : INTB * 3 / 4; int : INTB / 4; };
+struct V { int : INTB * 3 / 4; int j : INTB / 4; };
+struct W { int i; int : 0; int j; };
+static_assert (__has_unique_object_representations (U) == false, "");
+static_assert (__has_unique_object_representations (V) == false, "");
+static_assert (sizeof (W) != 2 * sizeof (int) || __has_unique_object_representations (W) == true, "");
Jakub
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: [PATCH] c++: Treat unnamed bitfields as padding for __has_unique_object_representations [PR109096]
2023-03-14 7:58 [PATCH] c++: Treat unnamed bitfields as padding for __has_unique_object_representations [PR109096] Jakub Jelinek
@ 2023-03-14 14:35 ` Jason Merrill
0 siblings, 0 replies; 2+ messages in thread
From: Jason Merrill @ 2023-03-14 14:35 UTC (permalink / raw)
To: Jakub Jelinek; +Cc: gcc-patches
On 3/14/23 03:58, Jakub Jelinek wrote:
> Hi!
>
> As reported in the PR, for __has_unique_object_representations we
> were treating unnamed bitfields as named ones, which is wrong, they
> are actually padding.
>
> THe following patch fixes that. Ok for trunk (and what about release
> branches later?)?
OK.
> 2023-03-14 Jakub Jelinek <jakub@redhat.com>
>
> PR c++/109096
> * tree.cc (record_has_unique_obj_representations): Ignore unnamed
> bitfields.
>
> * g++.dg/cpp1z/has-unique-obj-representations3.C: New test.
>
> --- gcc/cp/tree.cc.jj 2023-03-10 10:06:40.247560614 +0100
> +++ gcc/cp/tree.cc 2023-03-13 10:38:03.394836926 +0100
> @@ -4851,7 +4851,7 @@ record_has_unique_obj_representations (c
> DECL_SIZE (field)))
> return false;
> }
> - else if (DECL_C_BIT_FIELD (field))
> + else if (DECL_C_BIT_FIELD (field) && !DECL_UNNAMED_BIT_FIELD (field))
> {
> tree btype = DECL_BIT_FIELD_TYPE (field);
> if (!type_has_unique_obj_representations (btype))
> @@ -4862,7 +4862,7 @@ record_has_unique_obj_representations (c
>
> offset_int cur = 0;
> for (tree field = TYPE_FIELDS (t); field; field = DECL_CHAIN (field))
> - if (TREE_CODE (field) == FIELD_DECL)
> + if (TREE_CODE (field) == FIELD_DECL && !DECL_UNNAMED_BIT_FIELD (field))
> {
> offset_int fld = wi::to_offset (DECL_FIELD_OFFSET (field));
> offset_int bitpos = wi::to_offset (DECL_FIELD_BIT_OFFSET (field));
> --- gcc/testsuite/g++.dg/cpp1z/has-unique-obj-representations3.C.jj 2023-03-13 10:50:21.705127719 +0100
> +++ gcc/testsuite/g++.dg/cpp1z/has-unique-obj-representations3.C 2023-03-13 10:50:17.751185067 +0100
> @@ -0,0 +1,10 @@
> +// PR c++/109096
> +// { dg-do compile { target c++11 } }
> +
> +#define INTB (__SIZEOF_INT__ * __CHAR_BIT__)
> +struct U { int i : INTB * 3 / 4; int : INTB / 4; };
> +struct V { int : INTB * 3 / 4; int j : INTB / 4; };
> +struct W { int i; int : 0; int j; };
> +static_assert (__has_unique_object_representations (U) == false, "");
> +static_assert (__has_unique_object_representations (V) == false, "");
> +static_assert (sizeof (W) != 2 * sizeof (int) || __has_unique_object_representations (W) == true, "");
>
> Jakub
>
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2023-03-14 14:35 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2023-03-14 7:58 [PATCH] c++: Treat unnamed bitfields as padding for __has_unique_object_representations [PR109096] Jakub Jelinek
2023-03-14 14:35 ` Jason Merrill
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).