[Bug ipa/100413] New: [11/12 Regression] ICE: failed to reclaim unneeded function with custom flags

[Bug ipa/100413] New: [11/12 Regression] ICE: failed to reclaim unneeded function with custom flags

[zsojka at seznam dot cz]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100413

            Bug ID: 100413
           Summary: [11/12 Regression] ICE: failed to reclaim unneeded
                    function with custom flags
           Product: gcc
           Version: 12.0
            Status: UNCONFIRMED
          Keywords: ice-on-valid-code
          Severity: normal
          Priority: P3
         Component: ipa
          Assignee: unassigned at gcc dot gnu.org
          Reporter: zsojka at seznam dot cz
                CC: marxin at gcc dot gnu.org
  Target Milestone: ---
              Host: x86_64-pc-linux-gnu

Created attachment 50747
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=50747&action=edit
auto-reduced testcase (from OpenTTD sources)

Compiler output:
$ x86_64-pc-linux-gnu-gcc -O2 -fno-guess-branch-probability
-fno-inline-functions-called-once -fipa-cp-clone -fipa-pta
-fnon-call-exceptions --param=ipa-cp-eval-threshold=0 testcase.C 
_ZN12_Vector_base13_M_deallocateEv/6 (void _Vector_base::_M_deallocate())
@0x7fb58e54c880
  Type: function
  Body removed by symtab_remove_unreachable_nodes
  Visibility: externally_visible public weak comdat
comdat_group:_ZN12_Vector_base13_M_deallocateEv one_only
  References: 
  Referring: 
  Availability: not_available
  Function flags: body
  Called by: 
  Calls: 
testcase.C:60:73: internal compiler error: failed to reclaim unneeded function
   60 | ClientNetworkContentSocketHandler::ClientNetworkContentSocketHandler()
{}
      |                                                                        
^
0xe9f7ce mark_functions_to_output
        /repo/gcc-trunk/gcc/cgraphunit.c:1651
0xe9f7ce symbol_table::compile()
        /repo/gcc-trunk/gcc/cgraphunit.c:2321
0xea2007 symbol_table::compile()
        /repo/gcc-trunk/gcc/cgraphunit.c:2269
0xea2007 symbol_table::finalize_compilation_unit()
        /repo/gcc-trunk/gcc/cgraphunit.c:2537
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.

$ x86_64-pc-linux-gnu-gcc -v
Using built-in specs.
COLLECT_GCC=/repo/gcc-trunk/binary-latest/bin/x86_64-pc-linux-gnu-gcc
COLLECT_LTO_WRAPPER=/repo/gcc-trunk/binary-trunk-r12-429-20210504051736-g00d8545d705-checking-yes-rtl-df-extra-amd64/bin/../libexec/gcc/x86_64-pc-linux-gnu/12.0.0/lto-wrapper
Target: x86_64-pc-linux-gnu
Configured with: /repo/gcc-trunk//configure --enable-languages=c,c++
--enable-valgrind-annotations --disable-nls --enable-checking=yes,rtl,df,extra
--with-cloog --with-ppl --with-isl --build=x86_64-pc-linux-gnu
--host=x86_64-pc-linux-gnu --target=x86_64-pc-linux-gnu
--with-ld=/usr/bin/x86_64-pc-linux-gnu-ld
--with-as=/usr/bin/x86_64-pc-linux-gnu-as --disable-libstdcxx-pch
--prefix=/repo/gcc-trunk//binary-trunk-r12-429-20210504051736-g00d8545d705-checking-yes-rtl-df-extra-amd64
Thread model: posix
Supported LTO compression algorithms: zlib zstd
gcc version 12.0.0 20210504 (experimental) (GCC)

[Bug ipa/100413] [11/12 Regression] ICE: failed to reclaim unneeded function with custom flags

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100413

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |hubicka at gcc dot gnu.org
           Keywords|                            |ice-checking
   Target Milestone|---                         |11.2

[Bug ipa/100413] [11/12 Regression] ICE: failed to reclaim unneeded function with custom flags

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100413

Martin Liška <marxin at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|                            |2021-05-10
     Ever confirmed|0                           |1
           Assignee|unassigned at gcc dot gnu.org      |marxin at gcc dot gnu.org
             Status|UNCONFIRMED                 |ASSIGNED

--- Comment #1 from Martin Liška <marxin at gcc dot gnu.org> ---
I can take a look.

[Bug ipa/100413] [11/12 Regression] ICE: failed to reclaim unneeded function with custom flags

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100413

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|11.2                        |11.3

--- Comment #2 from Richard Biener <rguenth at gcc dot gnu.org> ---
GCC 11.2 is being released, retargeting bugs to GCC 11.3

[Bug ipa/100413] [11/12 Regression] ICE: failed to reclaim unneeded function with custom flags since r11-4267-g0e590b68fa374365

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100413

Martin Liška <marxin at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |mjambor at suse dot cz
           Assignee|marxin at gcc dot gnu.org          |unassigned at gcc dot gnu.org
            Summary|[11/12 Regression] ICE:     |[11/12 Regression] ICE:
                   |failed to reclaim unneeded  |failed to reclaim unneeded
                   |function with custom flags  |function with custom flags
                   |                            |since
                   |                            |r11-4267-g0e590b68fa374365
             Status|ASSIGNED                    |NEW

--- Comment #3 from Martin Liška <marxin at gcc dot gnu.org> ---
@Martin, can you please take a look? It's again something about clone
materialization starting with r11-4267-g0e590b68fa374365.

[Bug ipa/100413] [11/12 Regression] ICE: failed to reclaim unneeded function with custom flags since r11-4267-g0e590b68fa374365

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100413

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|2021-05-10 00:00:00         |2022-1-17
           Priority|P3                          |P2

--- Comment #4 from Richard Biener <rguenth at gcc dot gnu.org> ---
Re-confirmed.

[Bug ipa/100413] [11/12 Regression] ICE: failed to reclaim unneeded function with custom flags since r11-4267-g0e590b68fa374365

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100413

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|11.3                        |11.4

--- Comment #5 from Richard Biener <rguenth at gcc dot gnu.org> ---
GCC 11.3 is being released, retargeting bugs to GCC 11.4.

[Bug ipa/100413] [11/12 Regression] ICE: failed to reclaim unneeded function with custom flags since r11-4267-g0e590b68fa374365

[jamborm at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100413

Martin Jambor <jamborm at gcc dot gnu.org> changed:

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

--- Comment #6 from Martin Jambor <jamborm at gcc dot gnu.org> ---
I'm testing a patch.  The problem is that a clone is removed with
cgraph::remove and it leaves behind a dangling clone_of with a body for which
there is now no need.

[Bug ipa/100413] [11/12 Regression] ICE: failed to reclaim unneeded function with custom flags since r11-4267-g0e590b68fa374365

[jamborm at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100413

--- Comment #7 from Martin Jambor <jamborm at gcc dot gnu.org> ---
I proposed the aforementioned fix on the mailing list:
https://gcc.gnu.org/pipermail/gcc-patches/2022-April/593732.html

[Bug ipa/100413] [11/12/13 Regression] ICE: failed to reclaim unneeded function with custom flags since r11-4267-g0e590b68fa374365

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100413

--- Comment #8 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Martin Jambor <jamborm@gcc.gnu.org>:

https://gcc.gnu.org/g:27ee75dbe81bb781214c66a9e6a759c08b7deb60

commit r13-48-g27ee75dbe81bb781214c66a9e6a759c08b7deb60
Author: Martin Jambor <mjambor@suse.cz>
Date:   Fri Apr 29 17:38:15 2022 +0200

    ipa: Release body of clone_of when removing its last clone (PR 100413)

    In the PR, the verifier complains that we did not manage to remove the
    body of a node and it is right.  The node is kept for materialization
    of two clones but after one is materialized, the other one is removed
    as unneeded (as a part of delete_unreachable_blocks_update_callgraph).
    The problem is that the node removal does not check for this situation
    and can leave the clone_of node there with a body attached to it even
    though there is no use for it any more.  This patch does checks for it
    and handles the situation in a simlar way that
    cgraph_node::materialize_clone does it, except that it also has to be
    careful that the removed node itself does not have any clones, which
    would still need the clone_of's body.  Failing to do that results in a
    bootstrap failure.

    gcc/ChangeLog:

    2022-04-27  Martin Jambor  <mjambor@suse.cz>

            PR ipa/100413
            * cgraph.cc (cgraph_node::remove): Release body of the node this
            is clone_of if appropriate.

    gcc/testsuite/ChangeLog:

    2022-04-27  Martin Jambor  <mjambor@suse.cz>

            PR ipa/100413
            * g++.dg/ipa/pr100413.C: New test.

[Bug ipa/100413] [11/12/13 Regression] ICE: failed to reclaim unneeded function with custom flags since r11-4267-g0e590b68fa374365

[jamborm at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100413

--- Comment #9 from Martin Jambor <jamborm at gcc dot gnu.org> ---
Now fixed in trunk, I'll backport to 12 and 11 after 12.1 is released.

[Bug ipa/100413] [11/12/13 Regression] ICE: failed to reclaim unneeded function with custom flags since r11-4267-g0e590b68fa374365

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100413

--- Comment #10 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-12 branch has been updated by Martin Jambor
<jamborm@gcc.gnu.org>:

https://gcc.gnu.org/g:988fbc636639621dadc5cecbe77aa3f9aefedefb

commit r12-8368-g988fbc636639621dadc5cecbe77aa3f9aefedefb
Author: Martin Jambor <mjambor@suse.cz>
Date:   Wed May 11 23:25:40 2022 +0200

    ipa: Release body of clone_of when removing its last clone (PR 100413)

    In the PR, the verifier complains that we did not manage to remove the
    body of a node and it is right.  The node is kept for materialization
    of two clones but after one is materialized, the other one is removed
    as unneeded (as a part of delete_unreachable_blocks_update_callgraph).
    The problem is that the node removal does not check for this situation
    and can leave the clone_of node there with a body attached to it even
    though there is no use for it any more.  This patch does checks for it
    and handles the situation in a simlar way that
    cgraph_node::materialize_clone does it, except that it also has to be
    careful that the removed node itself does not have any clones, which
    would still need the clone_of's body.  Failing to do that results in a
    bootstrap failure.

    gcc/ChangeLog:

    2022-04-27  Martin Jambor  <mjambor@suse.cz>

            PR ipa/100413
            * cgraph.cc (cgraph_node::remove): Release body of the node this
            is clone_of if appropriate.

    gcc/testsuite/ChangeLog:

    2022-04-27  Martin Jambor  <mjambor@suse.cz>

            PR ipa/100413
            * g++.dg/ipa/pr100413.C: New test.

    (cherry picked from commit 27ee75dbe81bb781214c66a9e6a759c08b7deb60)

[Bug ipa/100413] [11/12/13 Regression] ICE: failed to reclaim unneeded function with custom flags since r11-4267-g0e590b68fa374365

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100413

--- Comment #11 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-11 branch has been updated by Martin Jambor
<jamborm@gcc.gnu.org>:

https://gcc.gnu.org/g:a03b46b78c82666b39ccada3df533f819e13dcb7

commit r11-10011-ga03b46b78c82666b39ccada3df533f819e13dcb7
Author: Martin Jambor <mjambor@suse.cz>
Date:   Thu May 19 14:06:08 2022 +0200

    ipa: Release body of clone_of when removing its last clone (PR 100413)

    In the PR, the verifier complains that we did not manage to remove the
    body of a node and it is right.  The node is kept for materialization
    of two clones but after one is materialized, the other one is removed
    as unneeded (as a part of delete_unreachable_blocks_update_callgraph).
    The problem is that the node removal does not check for this situation
    and can leave the clone_of node there with a body attached to it even
    though there is no use for it any more.  This patch does checks for it
    and handles the situation in a simlar way that
    cgraph_node::materialize_clone does it, except that it also has to be
    careful that the removed node itself does not have any clones, which
    would still need the clone_of's body.  Failing to do that results in a
    bootstrap failure.

    gcc/ChangeLog:

    2022-04-27  Martin Jambor  <mjambor@suse.cz>

            PR ipa/100413
            * cgraph.c (cgraph_node::remove): Release body of the node this
            is clone_of if appropriate.

    gcc/testsuite/ChangeLog:

    2022-04-27  Martin Jambor  <mjambor@suse.cz>

            PR ipa/100413
            * g++.dg/ipa/pr100413.C: New test.

    (cherry picked from commit 27ee75dbe81bb781214c66a9e6a759c08b7deb60)

[Bug ipa/100413] [11/12/13 Regression] ICE: failed to reclaim unneeded function with custom flags since r11-4267-g0e590b68fa374365

[jamborm at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100413

Martin Jambor <jamborm at gcc dot gnu.org> changed:

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

--- Comment #12 from Martin Jambor <jamborm at gcc dot gnu.org> ---
Fixed, thanks for reporting and sorry that it took so long to address.