[Bug c++/90448] [8/9/10 Regression] decltype-based lambda parameter pack is rejected

[Bug c++/90448] [8/9/10 Regression] decltype-based lambda parameter pack is rejected

[paolo.carlini at oracle dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90448

--- Comment #3 from Paolo Carlini <paolo.carlini at oracle dot com> ---
This is already fixed in trunk: I'm adding the testcase and removingg the 10
Regression marker.

[Bug c++/90448] [8/9/10 Regression] decltype-based lambda parameter pack is rejected

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90448

--- Comment #4 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Paolo Carlini <paolo@gcc.gnu.org>:

https://gcc.gnu.org/g:183e9006cc0300257d6f74d573514ef86f78cb4e

commit r10-7885-g183e9006cc0300257d6f74d573514ef86f78cb4e
Author: Paolo Carlini <paolo.carlini@oracle.com>
Date:   Wed Apr 22 18:05:18 2020 +0200

    Add testcase of PR c++/90448, already fixed in trunk.

            PR c++/90448
            * g++.dg/cpp1y/lambda-generic-variadic20.C: New.

[Bug c++/90448] [8/9 Regression] decltype-based lambda parameter pack is rejected

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90448

--- Comment #5 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
The testcase ICEs on powerpc64-linux with -m32:
lambda-generic-variadic20.C:5:12: internal compiler error: in
expand_expr_addr_expr_1, at expr.c:8075
    5 |   auto L = [](auto ... a) {
      |            ^~~~~~~~~~~~~~~~
    6 |     auto M = [](decltype(a) ... b) -> void {
      |     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
    7 |     };
      |     ~~      
    8 |     return M;
      |     ~~~~~~~~~
    9 |   };
      |   ~         
0xd9422a expand_expr_addr_expr_1
        ../../gcc/expr.c:8075
0xd94817 expand_expr_addr_expr
        ../../gcc/expr.c:8188
0xda0a68 expand_expr_real_1(tree_node*, rtx_def*, machine_mode,
expand_modifier, rtx_def**, bool)
        ../../gcc/expr.c:11363
0xd94bcd expand_expr_real(tree_node*, rtx_def*, machine_mode, expand_modifier,
rtx_def**, bool)
        ../../gcc/expr.c:8358
0xc050ff expand_normal
        ../../gcc/expr.h:288
0xc0686f precompute_register_parameters
        ../../gcc/calls.c:982
0xc0fc27 expand_call(tree_node*, rtx_def*, int)
        ../../gcc/calls.c:4398
0xd9f9cb expand_expr_real_1(tree_node*, rtx_def*, machine_mode,
expand_modifier, rtx_def**, bool)
        ../../gcc/expr.c:11135
0xd94bcd expand_expr_real(tree_node*, rtx_def*, machine_mode, expand_modifier,
rtx_def**, bool)
        ../../gcc/expr.c:8358
0xd8cef5 store_expr(tree_node*, rtx_def*, int, bool, bool)
        ../../gcc/expr.c:5752
0xd8c194 expand_assignment(tree_node*, tree_node*, bool)
        ../../gcc/expr.c:5514
0xc2860b expand_call_stmt
        ../../gcc/cfgexpand.c:2701
0xc2b6b5 expand_gimple_stmt_1
        ../../gcc/cfgexpand.c:3682
0xc2bc49 expand_gimple_stmt
        ../../gcc/cfgexpand.c:3847
0xc32a16 expand_gimple_basic_block
        ../../gcc/cfgexpand.c:5887
0xc3401a execute
        ../../gcc/cfgexpand.c:6542
Please submit a full bug report,
with preprocessed source if appropriate.
Please include the complete backtrace with any bug report.
See <https://gcc.gnu.org/bugs/> for instructions.

[Bug c++/90448] [8/9 Regression] decltype-based lambda parameter pack is rejected

[ro at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90448

Rainer Orth <ro at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |ro at gcc dot gnu.org

--- Comment #6 from Rainer Orth <ro at gcc dot gnu.org> ---
I see the same ICE on 32-bit sparc-sun-solaris2.11.

[Bug c++/90448] [8/9 Regression] decltype-based lambda parameter pack is rejected

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90448

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |hubicka at gcc dot gnu.org,
                   |                            |segher at gcc dot gnu.org

--- Comment #7 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
So, seems the second argument of the call is some empty class - struct
._anon_2, 1 byte long, and the FE in the CALL_EXPR passes an empty CONSTRUCTOR
of that type.
The gimplifier then creates a temporary for that (and doesn't initialize it),
so we have:
  struct ._anon_2 D.3223;

  <bb 2> :
  <retval> = fooV<{const char*, int, double, char, float, short int, unsigned
int}>::<lambda(auto:1 ...)>::operator()<fooV<{const char*, int, doubl
e, char, float, short int, unsigned int}>::<lambda(auto:1 ...)>, const char*,
int, double, char, float, short int, unsigned int> (0B, D.3223, _2(D)
, _3(D), _4(D), _5(D), _6(D), _7(D), _8(D)); [return slot optimization]
up to *.optimized in a thunk, D.3223 is not initialized (contains just padding)
and is passed by value to the call.
Now, on powerpc -m32, the ABI wants to pass that by reference -
if (pass_by_reference (args_so_far_pnt, arg))
is true, and we enter:
          /* If we're compiling a thunk, pass through invisible references
             instead of making a copy.  */
          if (call_from_thunk_p
              || (callee_copies
                  && !TREE_ADDRESSABLE (type)
                  && (base = get_base_address (args[i].tree_value))
                  && TREE_CODE (base) != SSA_NAME
                  && (!DECL_P (base) || MEM_P (DECL_RTL (base)))))
block because call_from_thunk_p is true.  But, base is a DECL with DECL_RTL of
a (reg:QI ...).
The code calls mark_addressable on args[i].tree_value, but that doesn't do
anything immediately, because currently_expanding_to_rtl is true (but even if
it would, that doesn't change DECL_RTL of the arg being passed).
Later the code calls build_fold_addr_expr_loc and ICEs when expanding that,
because trying to expand ADDR_EXPR on a VAR_DECL with DECL_RTL of (reg:QI ...)
doesn't work.

I don't see how this can be implemented other than actually making a copy,
so wonder if the condition shouldn't be
          if ((call_from_thunk_p || callee_copies)
              && !TREE_ADDRESSABLE (type)
              && ...
instead.  But what do I know about thunks.

[Bug c++/90448] [8/9/10/11 Regression] decltype-based lambda parameter pack is rejected

[ro at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90448

Rainer Orth <ro at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[8/9 Regression]            |[8/9/10/11 Regression]
                   |decltype-based lambda       |decltype-based lambda
                   |parameter pack is rejected  |parameter pack is rejected

--- Comment #8 from Rainer Orth <ro at gcc dot gnu.org> ---
At least on 32-bit Solaris/SPARC, the ICE also occurs on the gcc-10 branch and
master.

[Bug c++/90448] [8/9/10/11 Regression] decltype-based lambda parameter pack is rejected

[ro at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90448

Rainer Orth <ro at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jozefl at gcc dot gnu.org

--- Comment #9 from Rainer Orth <ro at gcc dot gnu.org> ---
*** Bug 96458 has been marked as a duplicate of this bug. ***

[Bug c++/90448] [8/9/10/11 Regression] decltype-based lambda parameter pack is rejected

[seurer at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90448

seurer at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |seurer at gcc dot gnu.org

--- Comment #10 from seurer at gcc dot gnu.org ---
*** Bug 99392 has been marked as a duplicate of this bug. ***

[Bug c++/90448] [8/9/10/11 Regression] decltype-based lambda parameter pack is rejected

[seurer at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90448

--- Comment #11 from seurer at gcc dot gnu.org ---
This (still) fails on powerpc BE 32 bit compiler for both gcc 10 and trunk.

[Bug c++/90448] [8/9/10/11 Regression] decltype-based lambda parameter pack is rejected

[ebotcazou at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90448

Eric Botcazou <ebotcazou at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED
           Assignee|unassigned at gcc dot gnu.org      |ebotcazou at gcc dot gnu.org
                 CC|                            |ebotcazou at gcc dot gnu.org

--- Comment #12 from Eric Botcazou <ebotcazou at gcc dot gnu.org> ---
Likewise on SPARC 32-bit, so let's find a way out.

[Bug c++/90448] [8/9/10/11 Regression] decltype-based lambda parameter pack is rejected

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90448

--- Comment #13 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Eric Botcazou <ebotcazou@gcc.gnu.org>:

https://gcc.gnu.org/g:defafb78cbcb7e2d5abd40379afdecee8b1a0d06

commit r11-7579-gdefafb78cbcb7e2d5abd40379afdecee8b1a0d06
Author: Eric Botcazou <ebotcazou@adacore.com>
Date:   Tue Mar 9 16:12:22 2021 +0100

    Fix internal error on lambda function

    This boils down to the RTL expander trying to take the address of a DECL
    whose RTX is a register.

    gcc/
            PR c++/90448
            * calls.c (initialize_argument_information): When the argument
            is passed by reference, do not make a copy in a thunk only if
            the argument is already in memory.  Remove redundant test for
            the case of callee copy.

[Bug c++/90448] [8/9/10/11 Regression] decltype-based lambda parameter pack is rejected

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90448

--- Comment #14 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-10 branch has been updated by Eric Botcazou
<ebotcazou@gcc.gnu.org>:

https://gcc.gnu.org/g:6051af8e3db9e6cea79d0e86da0e0dad286f7e4e

commit r10-9426-g6051af8e3db9e6cea79d0e86da0e0dad286f7e4e
Author: Eric Botcazou <ebotcazou@adacore.com>
Date:   Tue Mar 9 16:12:22 2021 +0100

    Fix internal error on lambda function

    This boils down to the RTL expander trying to take the address of a DECL
    whose RTX is a register.

    gcc/
            PR c++/90448
            * calls.c (initialize_argument_information): When the argument
            is passed by reference, do not make a copy in a thunk only if
            the argument is already in memory.  Remove redundant test for
            the case of callee copy.

[Bug c++/90448] [8/9/10/11 Regression] decltype-based lambda parameter pack is rejected

[ebotcazou at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90448

Eric Botcazou <ebotcazou at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
   Target Milestone|8.5                         |10.3
         Resolution|---                         |FIXED

--- Comment #15 from Eric Botcazou <ebotcazou at gcc dot gnu.org> ---
.

[Bug c++/90448] [8/9/10/11 Regression] decltype-based lambda parameter pack is rejected

[linkw at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90448

--- Comment #16 from Kewen Lin <linkw at gcc dot gnu.org> ---
*** Bug 90226 has been marked as a duplicate of this bug. ***