[Bug c/102505] New: [10/11/12 Regression] ICE in verify_sra_access_forest, at tree-sra.c:2368

[Bug c/102505] New: [10/11/12 Regression] ICE in verify_sra_access_forest, at tree-sra.c:2368

[gscfq@t-online.de]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102505

            Bug ID: 102505
           Summary: [10/11/12 Regression] ICE in verify_sra_access_forest,
                    at tree-sra.c:2368
           Product: gcc
           Version: 12.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c
          Assignee: unassigned at gcc dot gnu.org
          Reporter: gscfq@t-online.de
  Target Milestone: ---

Changed between 20200126 and 20200202, at -O2+ :
(case derived from g++.dg/cpp0x/nsdmi-virtual1a.C)
(gcc configured with --enable-checking=yes)


$ cat z1.cc
struct A
{
  A(): i(42) { }
  long double i;
  int f() { return i; }
};

struct D : A { int pad; };

struct B : virtual D
{
  int j = i + f();
  int k = A::i + A::f();
};

struct C: B { int pad; };

int main()
{
  C c;
  if (c.j != 84 || c.k != 84)
    __builtin_abort();
}


$ g++-12-20210926 -c z1.cc -O2
during GIMPLE pass: esra
z1.cc: In function 'int main()':
z1.cc:23:1: internal compiler error: in verify_sra_access_forest, at
tree-sra.c:2368
   23 | }
      | ^
0x1247974 verify_sra_access_forest(access*)
        ../../gcc/tree-sra.c:2367
0x1247c31 verify_all_sra_access_forests()
        ../../gcc/tree-sra.c:2434
0x124be95 analyze_all_variable_accesses
        ../../gcc/tree-sra.c:3497
0x124c916 perform_intra_sra
        ../../gcc/tree-sra.c:4673

[Bug c/102505] [10/11/12 Regression] ICE in verify_sra_access_forest, at tree-sra.c:2368

[gscfq@t-online.de]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102505

--- Comment #1 from G. Steinmetz <gscfq@t-online.de> ---

$ cat z2.cc
struct A
{
  A(): i(42) { }
  int i;
  int f() { return i; }
};

struct D : A { long double pad; };

struct B : virtual D
{
  int j = i + f();
  int k = A::i + A::f();
};

struct C: B { int pad; };

int main()
{
  C c;
  if (c.j != 84 || c.k != 84)
    __builtin_abort();
}

[Bug tree-optimization/102505] [10/11/12 Regression] ICE in verify_sra_access_forest, at tree-sra.c:2368

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102505

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jamborm at gcc dot gnu.org
   Last reconfirmed|                            |2021-09-27
     Ever confirmed|0                           |1
             Target|                            |x86_64-*-*
          Component|c++                         |tree-optimization
   Target Milestone|---                         |10.4
           See Also|                            |https://gcc.gnu.org/bugzill
                   |                            |a/show_bug.cgi?id=93516,
                   |                            |https://gcc.gnu.org/bugzill
                   |                            |a/show_bug.cgi?id=93667,
                   |                            |https://gcc.gnu.org/bugzill
                   |                            |a/show_bug.cgi?id=93845,
                   |                            |https://gcc.gnu.org/bugzill
                   |                            |a/show_bug.cgi?id=93776,
                   |                            |https://gcc.gnu.org/bugzill
                   |                            |a/show_bug.cgi?id=94598,
                   |                            |https://gcc.gnu.org/bugzill
                   |                            |a/show_bug.cgi?id=96730,
                   |                            |https://gcc.gnu.org/bugzill
                   |                            |a/show_bug.cgi?id=96820,
                   |                            |https://gcc.gnu.org/bugzill
                   |                            |a/show_bug.cgi?id=101626
             Status|UNCONFIRMED                 |NEW
           Keywords|                            |ice-on-valid-code

--- Comment #2 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Confirmed more reduced:
struct D  { int i;  int pad alignas(16); };
struct B : virtual D
{
  int j =84;
  int k =84;
};

struct C: B { };

int main()
{
  C c;
  if (c.j != 84 || c.k != 84)
    __builtin_abort();
}
---- CUT ----

Most likely  r10-6276-g1d8593070a62b.

What I don't get is why the ICE only happens on x86_64 and not on aarch64.  I
moved over to using alignas instead of long double to be similar on both
targets.

[Bug tree-optimization/102505] [10/11/12 Regression] ICE in verify_sra_access_forest with 16 byte aligned field and virtual inheritance

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102505

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Priority|P3                          |P2

[Bug tree-optimization/102505] [10/11/12 Regression] ICE in verify_sra_access_forest with 16 byte aligned field and virtual inheritance since r10-6321-g636e80eea24b780f

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102505

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[10/11/12 Regression] ICE   |[10/11/12 Regression] ICE
                   |in verify_sra_access_forest |in verify_sra_access_forest
                   |with 16 byte aligned field  |with 16 byte aligned field
                   |and virtual inheritance     |and virtual inheritance
                   |                            |since
                   |                            |r10-6321-g636e80eea24b780f
                 CC|                            |marxin at gcc dot gnu.org

--- Comment #3 from Martin Liška <marxin at gcc dot gnu.org> ---
No, it started with r10-6321-g636e80eea24b780f.

[Bug tree-optimization/102505] [10/11/12 Regression] ICE in verify_sra_access_forest with 16 byte aligned field and virtual inheritance since r10-6321-g636e80eea24b780f

[jamborm at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102505

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

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

--- Comment #4 from Martin Jambor <jamborm at gcc dot gnu.org> ---
Mine.

[Bug tree-optimization/102505] [10/11/12 Regression] ICE in verify_sra_access_forest with 16 byte aligned field and virtual inheritance since r10-6321-g636e80eea24b780f

[jamborm at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102505

--- Comment #5 from Martin Jambor <jamborm at gcc dot gnu.org> ---
I proposed a patch on the mailing list:
https://gcc.gnu.org/pipermail/gcc-patches/2021-October/582249.html

[Bug tree-optimization/102505] [10/11/12 Regression] ICE in verify_sra_access_forest with 16 byte aligned field and virtual inheritance since r10-6321-g636e80eea24b780f

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102505

--- Comment #6 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:701ee067807b80957c65bd7ff94b6099a27181de

commit r12-4607-g701ee067807b80957c65bd7ff94b6099a27181de
Author: Martin Jambor <mjambor@suse.cz>
Date:   Thu Oct 21 14:26:45 2021 +0200

    sra: Fix corner case of total scalarization with virtual inheritance (PR
102505)

    PR 102505 is a situation where of SRA takes its initial top-level
    access size from a get_ref_base_and_extent called on a COMPONENT_REF,
    and thus derived frm the FIELD_DECL, which however does not include a
    virtual base.  Total scalarization then goes on traversing the type,
    which however has virtual base past the non-virtual bits, tricking SRA
    to create sub-accesses outside of the supposedly encompassing
    accesses, which in turn triggers the verifier within the pass.

    The patch below fixes that by failing total scalarization when this
    situation is detected.

    gcc/ChangeLog:

    2021-10-20  Martin Jambor  <mjambor@suse.cz>

            PR tree-optimization/102505
            * tree-sra.c (totally_scalarize_subtree): Check that the
            encountered field fits within the acces we would like to put it
            in.

    gcc/testsuite/ChangeLog:

    2021-10-20  Martin Jambor  <mjambor@suse.cz>

            PR tree-optimization/102505
            * g++.dg/torture/pr102505.C: New test.

[Bug tree-optimization/102505] [10/11/12 Regression] ICE in verify_sra_access_forest with 16 byte aligned field and virtual inheritance since r10-6321-g636e80eea24b780f

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102505

--- Comment #7 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:f217e87972a2a207e793101fc05cfc9dd095c678

commit r12-4662-gf217e87972a2a207e793101fc05cfc9dd095c678
Author: Martin Jambor <mjambor@suse.cz>
Date:   Mon Oct 25 15:22:06 2021 +0200

    sra: Fix the fix for PR 102505 (PR 102886)

    I was not careful with the fix for PR 102505 and did not craft the
    check to satisfy the verifier carefully, which lead to PR 102886.
    (The verifier has the test structured differently and somewhat
    redundantly, so I could not just copy it).

    This patch fixes it.  I hope it is quite obvious correction of an
    oversight and so will commit it if survives bootstrap and testing on
    x86_64-linux and ppc64le-linux.

    Testcase for this bug is gcc.dg/tree-ssa/sra-18.c (but only on
    platforms with constant pools).  I will backport the two fixes
    to the release branches squashed.

    gcc/ChangeLog:

    2021-10-22  Martin Jambor  <mjambor@suse.cz>

            PR tree-optimization/102886
            * tree-sra.c (totally_scalarize_subtree): Fix the out of
            access-condition.

[Bug tree-optimization/102505] [10/11/12 Regression] ICE in verify_sra_access_forest with 16 byte aligned field and virtual inheritance since r10-6321-g636e80eea24b780f

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102505

--- Comment #8 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:97cebda80d367737c0b25909af8063b63166e00b

commit r11-9189-g97cebda80d367737c0b25909af8063b63166e00b
Author: Martin Jambor <mjambor@suse.cz>
Date:   Wed Oct 27 19:15:33 2021 +0200

    sra: Fix corner case of total scalarization with virtual inheritance (PR
102505)

    PR 102505 is a situation where of SRA takes its initial top-level
    access size from a get_ref_base_and_extent called on a COMPONENT_REF,
    and thus derived frm the FIELD_DECL, which however does not include a
    virtual base.  Total scalarization then goes on traversing the type,
    which however has virtual base past the non-virtual bits, tricking SRA
    to create sub-accesses outside of the supposedly encompassing
    accesses, which in turn triggers the verifier within the pass.

    The patch below fixes that by failing total scalarization when this
    situation is detected.

    This backport also has commit f217e87972a2a207e793101fc05cfc9dd095c678
    squashed into it in order to avoid PR 102886 that the fix introduced
    on trunk.

    gcc/ChangeLog:

    2021-10-20  Martin Jambor  <mjambor@suse.cz>

            PR tree-optimization/102505
            * tree-sra.c (totally_scalarize_subtree): Check that the
            encountered field fits within the acces we would like to put it
            in.

    gcc/testsuite/ChangeLog:

    2021-10-20  Martin Jambor  <mjambor@suse.cz>

            PR tree-optimization/102505
            * g++.dg/torture/pr102505.C: New test.

    (cherry picked from commit 701ee067807b80957c65bd7ff94b6099a27181de)

[Bug tree-optimization/102505] [10/11/12 Regression] ICE in verify_sra_access_forest with 16 byte aligned field and virtual inheritance since r10-6321-g636e80eea24b780f

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102505

--- Comment #9 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-10 branch has been updated by Martin Jambor
<jamborm@gcc.gnu.org>:

https://gcc.gnu.org/g:7d398e42d3d4ded84cb7a86396c75c5aaf518283

commit r10-10249-g7d398e42d3d4ded84cb7a86396c75c5aaf518283
Author: Martin Jambor <mjambor@suse.cz>
Date:   Tue Nov 2 11:17:51 2021 +0100

    sra: Fix corner case of total scalarization with virtual inheritance (PR
102505)

    PR 102505 is a situation where of SRA takes its initial top-level
    access size from a get_ref_base_and_extent called on a COMPONENT_REF,
    and thus derived frm the FIELD_DECL, which however does not include a
    virtual base.  Total scalarization then goes on traversing the type,
    which however has virtual base past the non-virtual bits, tricking SRA
    to create sub-accesses outside of the supposedly encompassing
    accesses, which in turn triggers the verifier within the pass.

    The patch below fixes that by failing total scalarization when this
    situation is detected.

    This backport also has commit f217e87972a2a207e793101fc05cfc9dd095c678
    squashed into it in order to avoid PR 102886 that the fix introduced
    on trunk.

    gcc/ChangeLog:

    2021-10-20  Martin Jambor  <mjambor@suse.cz>

            PR tree-optimization/102505
            * tree-sra.c (totally_scalarize_subtree): Check that the
            encountered field fits within the acces we would like to put it
            in.

    gcc/testsuite/ChangeLog:

    2021-10-20  Martin Jambor  <mjambor@suse.cz>

            PR tree-optimization/102505
            * g++.dg/torture/pr102505.C: New test.

    (cherry picked from commit 701ee067807b80957c65bd7ff94b6099a27181de)

[Bug tree-optimization/102505] [10/11/12 Regression] ICE in verify_sra_access_forest with 16 byte aligned field and virtual inheritance since r10-6321-g636e80eea24b780f

[jamborm at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102505

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

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

--- Comment #10 from Martin Jambor <jamborm at gcc dot gnu.org> ---
Fixed on master and the gcc-10 and gcc-11 release branches.