public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
* C++ PATCH for c++/70067 (ICE with typename typedef)
@ 2016-03-04 16:06 Jason Merrill
  2016-03-04 19:43 ` H.J. Lu
  0 siblings, 1 reply; 4+ messages in thread
From: Jason Merrill @ 2016-03-04 16:06 UTC (permalink / raw)
  To: gcc-patches List

[-- Attachment #1: Type: text/plain, Size: 281 bytes --]

On this testcase, when strip_typedefs rebuilds a TYPENAME_TYPE to remove 
the typedef, in this case it looks up the same typedef and we then abort 
because we still have a typedef.  In that case, strip the typedef 
explicitly.

Tested x86_64-pc-linux-gnu, applying to trunk and 5.

[-- Attachment #2: 70067.patch --]
[-- Type: text/x-patch, Size: 1213 bytes --]

commit 3f1401348184108dd03d3b0d76e51d5166ca95bf
Author: Jason Merrill <jason@redhat.com>
Date:   Thu Mar 3 22:40:40 2016 -0500

    	PR c++/70067
    	* tree.c (strip_typedefs): Handle TYPENAME_TYPE lookup finding the
    	same type.

diff --git a/gcc/cp/tree.c b/gcc/cp/tree.c
index 0b7b144..aaf9a4f 100644
--- a/gcc/cp/tree.c
+++ b/gcc/cp/tree.c
@@ -1437,6 +1437,9 @@ strip_typedefs (tree t, bool *remove_attributes)
 	result = make_typename_type (strip_typedefs (TYPE_CONTEXT (t),
 						     remove_attributes),
 				     fullname, typename_type, tf_none);
+	/* Handle 'typedef typename A::N N;'  */
+	if (typedef_variant_p (result))
+	  result = TYPE_MAIN_VARIANT (DECL_ORIGINAL_TYPE (TYPE_NAME (result)));
       }
       break;
     case DECLTYPE_TYPE:
diff --git a/gcc/testsuite/g++.dg/template/typename21.C b/gcc/testsuite/g++.dg/template/typename21.C
new file mode 100644
index 0000000..e5e59b1
--- /dev/null
+++ b/gcc/testsuite/g++.dg/template/typename21.C
@@ -0,0 +1,11 @@
+// PR c++/70067
+// { dg-do compile { target c++98 } }
+
+template <class> struct A;
+template <class T> struct B { struct N { }; };
+template <class T> struct D: B<T> {
+  typedef typename D::N N;
+  A<N> *a;
+};
+
+D<int> d;

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: C++ PATCH for c++/70067 (ICE with typename typedef)
  2016-03-04 16:06 C++ PATCH for c++/70067 (ICE with typename typedef) Jason Merrill
@ 2016-03-04 19:43 ` H.J. Lu
  2016-03-04 19:47   ` Jason Merrill
  0 siblings, 1 reply; 4+ messages in thread
From: H.J. Lu @ 2016-03-04 19:43 UTC (permalink / raw)
  To: Jason Merrill; +Cc: gcc-patches List

On Fri, Mar 4, 2016 at 8:06 AM, Jason Merrill <jason@redhat.com> wrote:
> On this testcase, when strip_typedefs rebuilds a TYPENAME_TYPE to remove the
> typedef, in this case it looks up the same typedef and we then abort because
> we still have a typedef.  In that case, strip the typedef explicitly.
>
> Tested x86_64-pc-linux-gnu, applying to trunk and 5.

I checked this into trunk for:

ERROR: g++.dg/template/typename21.C  -std=c++11: syntax error in
target selector "target c++98" for " dg-do 2 compile { target c++98 }
"
ERROR: g++.dg/template/typename21.C  -std=c++14: syntax error in
target selector "target c++98" for " dg-do 2 compile { target c++98 }
"
ERROR: g++.dg/template/typename21.C  -std=c++98: syntax error in
target selector "target c++98" for " dg-do 2 compile { target c++98 }
"

Will backport to 5 if needed.

-- 
H.J.
Index: ChangeLog
===================================================================
--- ChangeLog (revision 233974)
+++ ChangeLog (working copy)
@@ -1,3 +1,7 @@
+2016-03-04  H.J. Lu  <hongjiu.lu@intel.com>
+
+ * g++.dg/template/typename21.C: Replace c++98 with c++98_only.
+
 2016-03-04  David Malcolm  <dmalcolm@redhat.com>

  PR c/68187
Index: g++.dg/template/typename21.C
===================================================================
--- g++.dg/template/typename21.C (revision 233974)
+++ g++.dg/template/typename21.C (working copy)
@@ -1,5 +1,5 @@
 // PR c++/70067
-// { dg-do compile { target c++98 } }
+// { dg-do compile { target c++98_only } }

 template <class> struct A;
 template <class T> struct B { struct N { }; };

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: C++ PATCH for c++/70067 (ICE with typename typedef)
  2016-03-04 19:43 ` H.J. Lu
@ 2016-03-04 19:47   ` Jason Merrill
  2016-03-04 19:54     ` H.J. Lu
  0 siblings, 1 reply; 4+ messages in thread
From: Jason Merrill @ 2016-03-04 19:47 UTC (permalink / raw)
  To: H.J. Lu; +Cc: gcc-patches List

c++98_only is wrong, we should just remove the target specifier.

Jason

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: C++ PATCH for c++/70067 (ICE with typename typedef)
  2016-03-04 19:47   ` Jason Merrill
@ 2016-03-04 19:54     ` H.J. Lu
  0 siblings, 0 replies; 4+ messages in thread
From: H.J. Lu @ 2016-03-04 19:54 UTC (permalink / raw)
  To: Jason Merrill; +Cc: gcc-patches List

On Fri, Mar 4, 2016 at 11:47 AM, Jason Merrill <jason@redhat.com> wrote:
> c++98_only is wrong, we should just remove the target specifier.
>
> Jason

Tested with trunk and 5.  Done.

-- 
H.J.

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2016-03-04 19:54 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2016-03-04 16:06 C++ PATCH for c++/70067 (ICE with typename typedef) Jason Merrill
2016-03-04 19:43 ` H.J. Lu
2016-03-04 19:47   ` Jason Merrill
2016-03-04 19:54     ` H.J. Lu

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).