[Bug c++/111224] New: modules: xtreme-header-1_a.H etc. ICE (in core_vals, at cp/module.cc:6108) on AArch64

[Bug c++/111224] New: modules: xtreme-header-1_a.H etc. ICE (in core_vals, at cp/module.cc:6108) on AArch64

[xry111 at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111224

            Bug ID: 111224
           Summary: modules: xtreme-header-1_a.H etc. ICE (in core_vals,
                    at cp/module.cc:6108) on AArch64
           Product: gcc
           Version: 14.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: xry111 at gcc dot gnu.org
  Target Milestone: ---

FAIL: g++.dg/modules/xtreme-header-1_a.H -std=c++17 (internal compiler error:
in core_vals, at cp/module.cc:6108)
FAIL: g++.dg/modules/xtreme-header-1_a.H -std=c++17 (test for excess errors)
FAIL: g++.dg/modules/xtreme-header-1_a.H module-cmi 
(gcm.cache/$srcdir/g++.dg/modules/xtreme-header-1_a.H.gcm)
FAIL: g++.dg/modules/xtreme-header-1_b.C -std=c++17 (test for excess errors)
FAIL: g++.dg/modules/xtreme-header-1_c.C -std=c++17 (test for excess errors)
FAIL: g++.dg/modules/xtreme-header-1_a.H -std=c++2a (internal compiler error:
in core_vals, at cp/module.cc:6108)
FAIL: g++.dg/modules/xtreme-header-1_a.H -std=c++2a (test for excess errors)
FAIL: g++.dg/modules/xtreme-header-1_a.H module-cmi 
(gcm.cache/$srcdir/g++.dg/modules/xtreme-header-1_a.H.gcm)
FAIL: g++.dg/modules/xtreme-header-1_b.C -std=c++2a (test for excess errors)
FAIL: g++.dg/modules/xtreme-header-1_c.C -std=c++2a (test for excess errors)
FAIL: g++.dg/modules/xtreme-header-1_a.H -std=c++2b (internal compiler error:
in core_vals, at cp/module.cc:6108)
FAIL: g++.dg/modules/xtreme-header-1_a.H -std=c++2b (test for excess errors)
FAIL: g++.dg/modules/xtreme-header-1_a.H module-cmi 
(gcm.cache/$srcdir/g++.dg/modules/xtreme-header-1_a.H.gcm)
FAIL: g++.dg/modules/xtreme-header-1_b.C -std=c++2b (test for excess errors)
FAIL: g++.dg/modules/xtreme-header-1_c.C -std=c++2b (test for excess errors)

[Bug c++/111224] modules: xtreme-header-1_a.H etc. ICE (in core_vals, at cp/module.cc:6108) on AArch64

[xry111 at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111224

--- Comment #1 from Xi Ruoyao <xry111 at gcc dot gnu.org> ---
The stack trace in g++.log:

/home/xry111/git-repos/gcc/gcc/testsuite/g++.dg/modules/xtreme-header-1_a.H:
internal compiler error: in core_vals, at cp/module.cc:6108
0x9563a3 trees_out::core_vals(tree_node*)
        ../../gcc/gcc/cp/module.cc:6108
0x95a5af trees_out::tree_node_vals(tree_node*)
        ../../gcc/gcc/cp/module.cc:7218
0x95a5af trees_out::tree_value(tree_node*)
        ../../gcc/gcc/cp/module.cc:9083
0x954463 trees_out::tree_node(tree_node*)
        ../../gcc/gcc/cp/module.cc:9281
0x955423 trees_out::core_vals(tree_node*)
        ../../gcc/gcc/cp/module.cc:5998
0x958cbb trees_out::tree_node_vals(tree_node*)
        ../../gcc/gcc/cp/module.cc:7218
0x958cbb trees_out::fn_parms_init(tree_node*)
        ../../gcc/gcc/cp/module.cc:10189
0x952233 trees_out::decl_value(tree_node*, depset*)
        ../../gcc/gcc/cp/module.cc:7782
0x95d017 depset::hash::find_dependencies(module_state*)
        ../../gcc/gcc/cp/module.cc:13328
0x95df5b module_state::write_begin(elf_out*, cpp_reader*, module_state_config&,
unsigned int&)
        ../../gcc/gcc/cp/module.cc:17895
0x95f1ef finish_module_processing(cpp_reader*)
        ../../gcc/gcc/cp/module.cc:20237
0x8c8feb c_parse_final_cleanups()
        ../../gcc/gcc/cp/decl2.cc:5184
0xb95a0f c_common_parse_file()
        ../../gcc/gcc/c-family/c-opts.cc:1275

[Bug c++/111224] modules: xtreme-header-1_a.H etc. ICE (in core_vals, at cp/module.cc:6108) on AArch64

[xry111 at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111224

--- Comment #2 from Xi Ruoyao <xry111 at gcc dot gnu.org> ---
It seems related to Glibc version.

[Bug c++/111224] modules: xtreme-header-1_a.H etc. ICE (in core_vals, at cp/module.cc:6108) on AArch64

[xry111 at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111224

Xi Ruoyao <xry111 at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|needs-reduction             |ice-on-valid-code

--- Comment #3 from Xi Ruoyao <xry111 at gcc dot gnu.org> ---
Reduced:

$ cat t.ii
typedef __SVBool_t __sv_bool_t;
void _ZGVsMxv_sin(__sv_bool_t);
$ ./git-repos/gcc-build/gcc/cc1plus t.ii -fmodule-header
 creating:./t.ii
t.ii: internal compiler error: in core_vals, at cp/module.cc:6108
0x800a27 trees_out::core_vals(tree_node*)
        ../../gcc/gcc/cp/module.cc:6108
0x803c93 trees_out::tree_node_vals(tree_node*)
        ../../gcc/gcc/cp/module.cc:7216
0x803c93 trees_out::tree_value(tree_node*)
        ../../gcc/gcc/cp/module.cc:9081
0x7ff4d3 trees_out::tree_node(tree_node*)
        ../../gcc/gcc/cp/module.cc:9279
0x7ffe6f trees_out::core_vals(tree_node*)
        ../../gcc/gcc/cp/module.cc:5998
0x8019ef trees_out::tree_node_vals(tree_node*)
        ../../gcc/gcc/cp/module.cc:7216
0x8019ef trees_out::fn_parms_init(tree_node*)
        ../../gcc/gcc/cp/module.cc:10187
0x80421b trees_out::decl_value(tree_node*, depset*)
        ../../gcc/gcc/cp/module.cc:7780
0x804c67 depset::hash::find_dependencies(module_state*)
        ../../gcc/gcc/cp/module.cc:13326
0x80515f module_state::write_begin(elf_out*, cpp_reader*, module_state_config&,
unsigned int&)
        ../../gcc/gcc/cp/module.cc:17893
0x80619f finish_module_processing(cpp_reader*)
        ../../gcc/gcc/cp/module.cc:20235
0x7aaf6b c_parse_final_cleanups()
        ../../gcc/gcc/cp/decl2.cc:5183
0x966a73 c_common_parse_file()
        ../../gcc/gcc/c-family/c-opts.cc:1266
Please submit a full bug report, with preprocessed source (by using
-freport-bug).
Please include the complete backtrace with any bug report.
See <https://gcc.gnu.org/bugs/> for instructions.

[Bug c++/111224] modules: xtreme-header-1_a.H etc. ICE (in core_vals, at cp/module.cc:6108) on AArch64

[xry111 at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111224

Xi Ruoyao <xry111 at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
      Known to fail|                            |11.1.0, 11.4.0, 12.3.0

--- Comment #4 from Xi Ruoyao <xry111 at gcc dot gnu.org> ---
Not a regression.

[Bug c++/111224] modules: xtreme-header-1_a.H etc. ICE (in core_vals, at cp/module.cc:6108) on AArch64 with SVE types

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111224

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
                 CC|                            |pinskia at gcc dot gnu.org
     Ever confirmed|0                           |1
   Last reconfirmed|                            |2024-02-11

--- Comment #5 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Confirmed.

[Bug c++/111224] modules: xtreme-header-1_a.H etc. ICE (in core_vals, at cp/module.cc:6108) on AArch64 with SVE types

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111224

--- Comment #6 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
    case POLY_INT_CST:
      gcc_unreachable (); /* Not supported in C++.  */


That is definitely not true. POLY_INT_CST can be created via target builtins
even.

[Bug c++/111224] modules: xtreme-header-1_a.H etc. ICE (in core_vals, at cp/module.cc:6108) on AArch64 with SVE types

[nshead at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111224

Nathaniel Shead <nshead at gcc dot gnu.org> changed:

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

--- Comment #7 from Nathaniel Shead <nshead at gcc dot gnu.org> ---
Created attachment 57586
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=57586&action=edit
Untested patch to implement POLY_INT_CST in modules

Here's a potential fix for this issue. But I only have access to an x86_64
machine currently, so this is completely untested.

[Bug c++/111224] modules: xtreme-header-1_a.H etc. ICE (in core_vals, at cp/module.cc:6108) on AArch64 with SVE types

[ppalka at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111224

Patrick Palka <ppalka at gcc dot gnu.org> changed:

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

--- Comment #8 from Patrick Palka <ppalka at gcc dot gnu.org> ---
(In reply to Nathaniel Shead from comment #7)
> Created attachment 57586 [details]
> Untested patch to implement POLY_INT_CST in modules
> 
> Here's a potential fix for this issue. But I only have access to an x86_64
> machine currently, so this is completely untested.
The compile farm might have a suitable machine for testing:
https://gcc.gnu.org/wiki/CompileFarm

[Bug c++/111224] modules: xtreme-header-1_a.H etc. ICE (in core_vals, at cp/module.cc:6108) on AArch64 with SVE types

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111224

--- Comment #9 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
(In reply to Nathaniel Shead from comment #7)
> Created attachment 57586 [details]
> Untested patch to implement POLY_INT_CST in modules
> 
> Here's a potential fix for this issue. But I only have access to an x86_64
> machine currently, so this is completely untested.

I can test it but it won't be until the middle of this up comming week.

[Bug c++/111224] modules: xtreme-header-1_a.H etc. ICE (in core_vals, at cp/module.cc:6108) on AArch64 with SVE types

[nshead at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111224

Nathaniel Shead <nshead at gcc dot gnu.org> changed:

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

--- Comment #10 from Nathaniel Shead <nshead at gcc dot gnu.org> ---
I now have access to an Aarch64 machine via the build farm. My patch in #c7 is
not sufficient to get the minimised testcase working (need to update the
assumptions on VECTOR_TYPEs too) and it looks like we'll need to work out how
to stream the aarch64-specific types found in the 'registered_builtin_types'
list too, like __fp16.

I'll submit a patch to the mailing lists once I've worked out the details.

[Bug c++/111224] modules: xtreme-header-1_a.H etc. ICE (in core_vals, at cp/module.cc:6108) on AArch64 with SVE types

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111224

--- Comment #11 from GCC Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Nathaniel Shead <nshead@gcc.gnu.org>:

https://gcc.gnu.org/g:4aa87b856067d4911de8fb66b3a27659dc75ca6d

commit r14-9439-g4aa87b856067d4911de8fb66b3a27659dc75ca6d
Author: Nathaniel Shead <nathanieloshead@gmail.com>
Date:   Sun Mar 10 22:06:18 2024 +1100

    c++: Support target-specific nodes when streaming modules [PR111224]

    Some targets make use of POLY_INT_CSTs and other custom builtin types,
    which currently violate some assumptions when streaming. This patch adds
    support for them, such as types like Aarch64 __fp16, PowerPC __ibm128,
    and vector types thereof.

    This patch doesn't provide "full" support of AArch64 SVE, however, since
    for that we would need to support 'target' nodes (tracked in PR108080).

    Adding the new builtin types means that on Aarch64 we now have 217
    global trees created on initialisation (up from 191), so this patch also
    slightly bumps the initial size of the fixed_trees allocation to 250.

            PR c++/98645
            PR c++/98688
            PR c++/111224

    gcc/cp/ChangeLog:

            * module.cc (enum tree_tag): Add new tag for builtin types.
            (trees_out::start): POLY_INT_CSTs can be emitted.
            (trees_in::start): Likewise.
            (trees_out::core_vals): Stream POLY_INT_CSTs.
            (trees_in::core_vals): Likewise.
            (trees_out::type_node): Handle vectors with multiple coeffs.
            (trees_in::tree_node): Likewise.
            (init_modules): Register target-specific builtin types. Bump
            initial capacity slightly.

    gcc/testsuite/ChangeLog:

            * g++.dg/modules/target-aarch64-1_a.C: New test.
            * g++.dg/modules/target-aarch64-1_b.C: New test.
            * g++.dg/modules/target-powerpc-1_a.C: New test.
            * g++.dg/modules/target-powerpc-1_b.C: New test.
            * g++.dg/modules/target-powerpc-2_a.C: New test.
            * g++.dg/modules/target-powerpc-2_b.C: New test.

    Signed-off-by: Nathaniel Shead <nathanieloshead@gmail.com>
    Reviewed-by: Patrick Palka <ppalka@redhat.com>