[Bug c++/101689] New: Internal compiler error during GIMPLE pass: ehcleanup

[Bug c++/101689] New: Internal compiler error during GIMPLE pass: ehcleanup

[Theodore.Papadopoulo at inria dot fr]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101689

            Bug ID: 101689
           Summary: Internal compiler error during GIMPLE pass: ehcleanup
           Product: gcc
           Version: 11.1.1
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: Theodore.Papadopoulo at inria dot fr
  Target Milestone: ---

Description of problem: The compiler crashes with the mentioned error.


Version-Release number of selected component (if applicable):
gcc version 11.1.1 20210531 (Red Hat 11.1.1-3) (GCC)

How reproducible:
Just compile the attached preprocessed code with the options given below.

Steps to Reproduce:
1. Download the attached code and name it ovp_main.i
2. /usr/bin/c++ -fvtable-verify=std -fnon-call-exceptions -std=gnu++11
ovp_main.i

Actual results:
mururoa-> /usr/bin/c++ -fvtable-verify=std -fnon-call-exceptions -std=gnu++11
ovp_main.i
during GIMPLE pass: ehcleanup
In file included from
/home/papadop/compiles/OpenViBe/OpenViBE-f34/dist/sdk-Debug/include/toolkit/ovtk_all.h:75,
                 from
/home/papadop/src/Git/OpenViBe/designer/plugins/examples/src/box-algorithms/ovpCBoxAlgorithmModifiableSettings.h:6,
                 from
/home/papadop/src/Git/OpenViBe/designer/plugins/examples/src/ovp_main.cpp:3:
/home/papadop/compiles/OpenViBe/OpenViBE-f34/dist/sdk-Debug/include/toolkit/box-algorithms/ovtkTBoxAlgorithm.h:
In member function ‘uint64_t
OpenViBE::Toolkit::TBoxAlgorithm<TBoxAlgorithmParentClass>::getClockFrequency(OpenViBE::Kernel::IBoxAlgorithmContext&)
[with TBoxAlgorithmParentClass = OpenViBE::Plugins::IBoxAlgorithm]’:
/home/papadop/compiles/OpenViBe/OpenViBE-f34/dist/sdk-Debug/include/toolkit/box-algorithms/ovtkTBoxAlgorithm.h:19:26:
internal compiler error: in mark_reachable_handlers, at tree-eh.c:4033
   19 |         virtual uint64_t
getClockFrequency(Kernel::IBoxAlgorithmContext& ctx)
      |                          ^~~~~~~~~~~~~~~~~
0x7f0bdc834b74 ???
        ../csu/libc-start.c:332
Please submit a full bug report,


Expected results:
No crash

[Bug c++/101689] Internal compiler error during GIMPLE pass: ehcleanup

[Theodore.Papadopoulo at inria dot fr]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101689

--- Comment #1 from Theodore.Papadopoulo at inria dot fr ---
Also submitted as 
Bug 1986538 - Internal compiler error during GIMPLE pass: ehcleanup
in readhat bugzilla, but I guess it is better here. Will close that one.

[Bug c++/101689] Internal compiler error during GIMPLE pass: ehcleanup

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101689

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

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

--- Comment #2 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
I don't see any attachments.  You can gzip/bzip2/xz compress it if it is too
large.

[Bug c++/101689] Internal compiler error during GIMPLE pass: ehcleanup

[Theodore.Papadopoulo at inria dot fr]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101689

--- Comment #3 from Theodore.Papadopoulo at inria dot fr ---
Created attachment 51224
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=51224&action=edit
Preprocessed code

[Bug c++/101689] Internal compiler error during GIMPLE pass: ehcleanup

[Theodore.Papadopoulo at inria dot fr]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101689

--- Comment #4 from Theodore.Papadopoulo at inria dot fr ---
Sorry, it was indeed too big, but I had not seen the message. Attached now.
Sorry for not having found time to reduce the test case.

[Bug c++/101689] Internal compiler error during GIMPLE pass: ehcleanup

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101689

--- Comment #5 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Reduced testcase:
struct A {
  virtual void foo (int x) { C e(&f, &x); bar (); }
  virtual bool bar ();
  struct C { C (int *, int *); };
  int f;
};
struct B : A {} d;

[Bug c++/101689] ICE with -fvtable-verify=std and -fnon-call-exceptions

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101689

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|Internal compiler error     |ICE with
                   |during GIMPLE pass:         |-fvtable-verify=std and
                   |ehcleanup                   |-fnon-call-exceptions
     Ever confirmed|0                           |1
   Last reconfirmed|                            |2021-07-30
             Status|UNCONFIRMED                 |NEW
           Keywords|                            |ice-on-valid-code

--- Comment #6 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Confirmed, this has been a bug since -fvtable-verify=std was introduced.
ICE with checking turned on:
<source>: In member function 'virtual void A::foo(int)':
<source>:2:16: error: statement marked for throw in middle of block
    2 |   virtual void foo (int x) { C e(&f, &x); bar (); }
      |                ^~~
# VUSE <.MEM_8>
_9 = this_6(D)->_vptr.A;
during GIMPLE pass: vtable-verify

[Bug c++/101689] ICE with -fvtable-verify=std and -fnon-call-exceptions

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101689

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |DUPLICATE
             Status|NEW                         |RESOLVED

--- Comment #7 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Dup of bug 59507.

*** This bug has been marked as a duplicate of bug 59507 ***