[Bug tree-optimization/94211] New: [9/10 Regression] -fcompare-debug failure on phi-opt-13.c

[Bug tree-optimization/94211] New: [9/10 Regression] -fcompare-debug failure on phi-opt-13.c

[zhroma at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94211

            Bug ID: 94211
           Summary: [9/10 Regression] -fcompare-debug failure on
                    phi-opt-13.c
           Product: gcc
           Version: 9.3.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: tree-optimization
          Assignee: unassigned at gcc dot gnu.org
          Reporter: zhroma at gcc dot gnu.org
  Target Milestone: ---

This command:
gcc -c -O2 -fcompare-debug gcc/testsuite/gcc.dg/tree-ssa/phi-opt-13.c

fails on todays master (r10-7237) on ppc64le, fails on 9.3 on x86_64, but not
in 7.4.  Haven't checked version 8.x.

long f(long a, long b) {
  if (__builtin_expect(b == 1, 1)) return a;
  return a / b;
}

Tree .phiopt2 is the first dump which seems buggy.

[Bug tree-optimization/94211] [9/10 Regression] -fcompare-debug failure on phi-opt-13.c

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94211

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
      Known to fail|                            |9.3.0
   Target Milestone|---                         |9.4
      Known to work|                            |7.4.0

[Bug tree-optimization/94211] [9/10 Regression] -fcompare-debug failure on phi-opt-13.c

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94211

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|                            |2020-03-18
     Ever confirmed|0                           |1
             Status|UNCONFIRMED                 |NEW
                 CC|                            |jakub at gcc dot gnu.org

--- Comment #1 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Started with r8-5241-g8697bf9f46f36168ddba5752db582e673e3cbe8c .  I'll have a
quick look.

[Bug tree-optimization/94211] [8/9/10 Regression] -fcompare-debug failure on phi-opt-13.c

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94211

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED
           Assignee|unassigned at gcc dot gnu.org      |jakub at gcc dot gnu.org
   Target Milestone|9.4                         |8.5
            Summary|[9/10 Regression]           |[8/9/10 Regression]
                   |-fcompare-debug failure on  |-fcompare-debug failure on
                   |phi-opt-13.c                |phi-opt-13.c

--- Comment #2 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
I think I've introduced this with
r8-3866-g14745bcac0e7b90a5c671b1f9402a53e57ea6431
With
  int e = a + 1;
before return a / b; it started already in that revision.

[Bug tree-optimization/94211] [8/9/10 Regression] -fcompare-debug failure on phi-opt-13.c

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94211

--- Comment #3 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Created attachment 48056
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=48056&action=edit
gcc10-pr94211.patch

Untested fix.

[Bug tree-optimization/94211] [8/9/10 Regression] -fcompare-debug failure on phi-opt-13.c

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94211

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

https://gcc.gnu.org/g:c7e9019681857b329bbe4c1e7ec8dec8c736c0fe

commit r10-7274-gc7e9019681857b329bbe4c1e7ec8dec8c736c0fe
Author: Jakub Jelinek <jakub@redhat.com>
Date:   Thu Mar 19 10:24:16 2020 +0100

    phiopt: Avoid -fcompare-debug bug in phiopt [PR94211]

    Two years ago, I've added support for up to 2 simple preparation statements
    in value_replacement, but the
    -      && estimate_num_insns (assign, &eni_time_weights)
    +      && estimate_num_insns (bb_seq (middle_bb), &eni_time_weights)
    change, meant that we compute the cost of all those statements rather than
    just the single assign that has been the single supported non-debug
    statement in the bb before, doesn't do what I thought would do, gimple_seq
    is just gimple * and thus it can't be really overloaded depending on
whether
    we pass a single gimple * or a whole sequence.  Which means in the last
    two years it doesn't count all the statements, but only the first one.
    With -g that happens to be a DEBUG_STMT, or it could be e.g. the first
    preparation statement which could be much cheaper than the actual assign.

    2020-03-19  Jakub Jelinek  <jakub@redhat.com>

            PR tree-optimization/94211
            * tree-ssa-phiopt.c (value_replacement): Use estimate_num_insns_seq
            instead of estimate_num_insns for bb_seq (middle_bb).  Rename
            emtpy_or_with_defined_p variable to empty_or_with_defined_p, adjust
            all uses.

            * gcc.dg/pr94211.c: New test.

[Bug tree-optimization/94211] [8/9 Regression] -fcompare-debug failure on phi-opt-13.c

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94211

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[8/9/10 Regression]         |[8/9 Regression]
                   |-fcompare-debug failure on  |-fcompare-debug failure on
                   |phi-opt-13.c                |phi-opt-13.c

--- Comment #5 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Fixed on the trunk so far.

[Bug tree-optimization/94211] [8/9 Regression] -fcompare-debug failure on phi-opt-13.c

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94211

--- Comment #6 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-9 branch has been updated by Jakub Jelinek
<jakub@gcc.gnu.org>:

https://gcc.gnu.org/g:8db876e9c045c57d2dc5bd08a6e250f822efaad0

commit r9-8464-g8db876e9c045c57d2dc5bd08a6e250f822efaad0
Author: Jakub Jelinek <jakub@redhat.com>
Date:   Thu Mar 19 10:24:16 2020 +0100

    phiopt: Avoid -fcompare-debug bug in phiopt [PR94211]

    Two years ago, I've added support for up to 2 simple preparation statements
    in value_replacement, but the
    -      && estimate_num_insns (assign, &eni_time_weights)
    +      && estimate_num_insns (bb_seq (middle_bb), &eni_time_weights)
    change, meant that we compute the cost of all those statements rather than
    just the single assign that has been the single supported non-debug
    statement in the bb before, doesn't do what I thought would do, gimple_seq
    is just gimple * and thus it can't be really overloaded depending on
whether
    we pass a single gimple * or a whole sequence.  Which means in the last
    two years it doesn't count all the statements, but only the first one.
    With -g that happens to be a DEBUG_STMT, or it could be e.g. the first
    preparation statement which could be much cheaper than the actual assign.

    2020-03-19  Jakub Jelinek  <jakub@redhat.com>

            PR tree-optimization/94211
            * tree-ssa-phiopt.c (value_replacement): Use estimate_num_insns_seq
            instead of estimate_num_insns for bb_seq (middle_bb).  Rename
            emtpy_or_with_defined_p variable to empty_or_with_defined_p, adjust
            all uses.

            * gcc.dg/pr94211.c: New test.

[Bug tree-optimization/94211] [8 Regression] -fcompare-debug failure on phi-opt-13.c

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94211

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[8/9 Regression]            |[8 Regression]
                   |-fcompare-debug failure on  |-fcompare-debug failure on
                   |phi-opt-13.c                |phi-opt-13.c

--- Comment #7 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Fixed for 9.4+ too.

[Bug tree-optimization/94211] [8 Regression] -fcompare-debug failure on phi-opt-13.c

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94211

--- Comment #8 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-8 branch has been updated by Jakub Jelinek
<jakub@gcc.gnu.org>:

https://gcc.gnu.org/g:2b0aaa4970a8d174b00fb98b4d24550d1c9a4e3a

commit r8-10471-g2b0aaa4970a8d174b00fb98b4d24550d1c9a4e3a
Author: Jakub Jelinek <jakub@redhat.com>
Date:   Thu Mar 19 10:24:16 2020 +0100

    phiopt: Avoid -fcompare-debug bug in phiopt [PR94211]

    Two years ago, I've added support for up to 2 simple preparation statements
    in value_replacement, but the
    -      && estimate_num_insns (assign, &eni_time_weights)
    +      && estimate_num_insns (bb_seq (middle_bb), &eni_time_weights)
    change, meant that we compute the cost of all those statements rather than
    just the single assign that has been the single supported non-debug
    statement in the bb before, doesn't do what I thought would do, gimple_seq
    is just gimple * and thus it can't be really overloaded depending on
whether
    we pass a single gimple * or a whole sequence.  Which means in the last
    two years it doesn't count all the statements, but only the first one.
    With -g that happens to be a DEBUG_STMT, or it could be e.g. the first
    preparation statement which could be much cheaper than the actual assign.

    2020-03-19  Jakub Jelinek  <jakub@redhat.com>

            PR tree-optimization/94211
            * tree-ssa-phiopt.c (value_replacement): Use estimate_num_insns_seq
            instead of estimate_num_insns for bb_seq (middle_bb).  Rename
            emtpy_or_with_defined_p variable to empty_or_with_defined_p, adjust
            all uses.

            * gcc.dg/pr94211.c: New test.

    (cherry picked from commit 8db876e9c045c57d2dc5bd08a6e250f822efaad0)

[Bug tree-optimization/94211] [8 Regression] -fcompare-debug failure on phi-opt-13.c

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94211

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

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

--- Comment #9 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Fixed for 8.5 too.