[Bug c++/94827] New: [ICE] [regression] crash on requires clause in tparam list

[Bug c++/94827] New: [ICE] [regression] crash on requires clause in tparam list

[eric.niebler at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94827

            Bug ID: 94827
           Summary: [ICE] [regression] crash on requires clause in tparam
                    list
           Product: gcc
           Version: 10.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: eric.niebler at gmail dot com
  Target Milestone: ---

Try the following code with -std=c++2a

template <typename T,
  bool = requires { requires (sizeof(T)==0); } >
void foo(T) {}

The result:

<source>:20:44: internal compiler error: tree check: accessed elt 1 of
'tree_vec' with 0 elts in map_arguments, at cp/constraint.cc:553
   20 |   bool = requires { requires (sizeof(T)==0); } >
      |                                            ^
Please submit a full bug report,
with preprocessed source if appropriate.
See <https://gcc.gnu.org/bugs/> for instructions.
Compiler returned: 1

This compiles with gcc-9 with -std=c++2a -fconcepts, so this is a regression.

[Bug c++/94827] [10 Regression] crash on requires clause in tparam list

[nathan at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94827

Nathan Sidwell <nathan at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |ASSIGNED
     Ever confirmed|0                           |1
   Last reconfirmed|                            |2020-04-28

--- Comment #1 from Nathan Sidwell <nathan at gcc dot gnu.org> ---
hm, yeah, we're completely unprepared to go inspecting the parm vector during
parsing the parameters.  Not sure how this worked previously.

[Bug c++/94827] [10 Regression] crash on requires clause in tparam list

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94827

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |ice-on-valid-code
           Priority|P3                          |P1
      Known to work|                            |9.3.0
   Target Milestone|---                         |10.0

[Bug c++/94827] [10 Regression] crash on requires clause in tparam list since r10-4424

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94827

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jakub at gcc dot gnu.org
            Summary|[10 Regression] crash on    |[10 Regression] crash on
                   |requires clause in tparam   |requires clause in tparam
                   |list                        |list since r10-4424

--- Comment #2 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Started with r10-4424-g81a34a6b68184436726489b81d44267c40f6fbe7

[Bug c++/94827] [10 Regression] crash on requires clause in tparam list since r10-4424

[nathan at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94827

--- Comment #3 from Nathan Sidwell <nathan at gcc dot gnu.org> ---
Created attachment 48425
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=48425&action=edit
try instantiating the fn

This one ices with the initial patch.

[Bug c++/94827] [10 Regression] crash on requires clause in tparam list since r10-4424

[nathan at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94827

--- Comment #4 from Nathan Sidwell <nathan at gcc dot gnu.org> ---
Note to stage-1 me:

Jason wrote:
But I don't think that we need to keep saving the converted
current_template_parms; diagnostics could also normalize using NULL_TREE args. 
And it looks like diagnose_nested_requirement isn't currently doing
re-normalization anyway.  This doesn't need to block the release, though.

[Bug c++/94827] crash on requires clause in tparam list since r10-4424

[nathan at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94827

Nathan Sidwell <nathan at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[10 Regression] crash on    |crash on requires clause in
                   |requires clause in tparam   |tparam list since r10-4424
                   |list since r10-4424         |

--- Comment #5 from Nathan Sidwell <nathan at gcc dot gnu.org> ---
fixed c416c52bcdb, leaving open for stage-1 issue

[Bug c++/94827] crash on requires clause in tparam list since r10-4424

[nathan at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94827

Nathan Sidwell <nathan at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|---                         |FIXED

--- Comment #6 from Nathan Sidwell <nathan at gcc dot gnu.org> ---
Cleanup pushed f1621d18f55