[Bug c++/114297] New: Yet more problems with "definition in block does not dominate use in block"

[Bug c++/114297] New: Yet more problems with "definition in block does not dominate use in block"

[dcb314 at hotmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114297

            Bug ID: 114297
           Summary: Yet more problems with "definition in block does not
                    dominate use in block"
           Product: gcc
           Version: 14.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: dcb314 at hotmail dot com
  Target Milestone: ---

For this C++ code:

void __assert_fail(char *, int, const char *) __attribute__((__noreturn__));
template <class T> void max(T, T);
template <class T> struct SimpleVector {
  T array;
  int num;
  T *getRef(int i) {
    num - i ? void() : __assert_fail("", 7, __PRETTY_FUNCTION__);
    return &array + i;
  }
};
struct Extremes {
  int minWidth;
  int maxWidth;
};
int forceCalcColumnExtremes_cs;
struct Table {
  SimpleVector<Extremes> colExtremes;
  void forceCalcColumnExtremes();
};
void Table::forceCalcColumnExtremes() {
  int minSumCols, maxSumCols;
  for (int i; i; ++i) {
    minSumCols += colExtremes.getRef(i)->minWidth;
    maxSumCols += colExtremes.getRef(i)->maxWidth;
  }
  max(forceCalcColumnExtremes_cs, minSumCols);
  max(forceCalcColumnExtremes_cs, maxSumCols);
}

compiled by recent gcc trunk, does this:

cvise $ ~/gcc/results/bin/g++ -c -w -O3 bug1022.cc
cvise $ ~/gcc/results/bin/g++ -c -w -O3 -march=znver3 bug1022.cc
bug1022.cc: In member function ‘void Table::forceCalcColumnExtremes()’:
bug1022.cc:20:6: error: definition in block 5 does not dominate use in block 3
   20 | void Table::forceCalcColumnExtremes() {
      |      ^~~~~
for SSA_NAME: vect_maxSumCols_16.16_76 in statement:
vect_maxSumCols_16.16_90 = PHI <vect_maxSumCols_16.16_76(3)>
PHI argument
vect_maxSumCols_16.16_76
for PHI node
vect_maxSumCols_16.16_90 = PHI <vect_maxSumCols_16.16_76(3)>
during GIMPLE pass: vect
bug1022.cc:20:6: internal compiler error: verify_ssa failed
0x159e432 verify_ssa(bool, bool)
        ../../trunk.20210101/gcc/tree-ssa.cc:1203
0x1204a9d execute_function_todo
        ../../trunk.20210101/gcc/passes.cc:2095

The bug first seems to occur sometime between g:71244316cf714725
and g:10cbfcd60f9e5bdb, which is a distance of 43 commits.

[Bug tree-optimization/114297] [14 Regression] Yet more problems with "definition in block does not dominate use in block"

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114297

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|Yet more problems with      |[14 Regression] Yet more
                   |"definition in block does   |problems with "definition
                   |not dominate use in block"  |in block does not dominate
                   |                            |use in block"
   Target Milestone|---                         |14.0
             Target|                            |x86_64

[Bug tree-optimization/114297] [14 Regression] Yet more problems with "definition in block does not dominate use in block"

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114297

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
     Ever confirmed|0                           |1
   Last reconfirmed|                            |2024-03-10
             Status|UNCONFIRMED                 |NEW
             Target|x86_64                      |x86_64 aarch64

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Reduced testcase:
```

void h() __attribute__((__noreturn__));
struct Extremes {
  int w;
  int h;
};
struct Extremes *array;
int f(int num, int size1)
{
  int sw = 0, sh = 0;
  for (int i = 0; i < size1; ++i)
  {
    if (num - i == 0)
      h();
    sw += array[i].w;
    sh += array[i].h;
  }
  return (sw) +  (sh);
}
```

Also ICEs on aarch64 with just -O3.

[Bug tree-optimization/114297] [14 Regression] Yet more problems with "definition in block does not dominate use in block"

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114297

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

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

--- Comment #2 from Richard Biener <rguenth at gcc dot gnu.org> ---
Let me have a look.

[Bug tree-optimization/114297] [14 Regression] Yet more problems with "definition in block does not dominate use in block"

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114297

--- Comment #3 from GCC Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Richard Biener <rguenth@gcc.gnu.org>:

https://gcc.gnu.org/g:c0c57246d5b47459bdb488734bc2c004a92668b5

commit r14-9435-gc0c57246d5b47459bdb488734bc2c004a92668b5
Author: Richard Biener <rguenther@suse.de>
Date:   Mon Mar 11 14:58:57 2024 +0100

    tree-optimization/114297 - SLP reduction with early break fix

    The following makes sure to pass in the SLP node for the live stmts
    we are generating the reduction epilogue for to
    vect_create_epilog_for_reduction.  This follows the previous fix for
    the non-SLP path.

            PR tree-optimization/114297
            * tree-vect-loop.cc (vectorizable_live_operation): Pass in the
            live stmts SLP node to vect_create_epilog_for_reduction.

            * gcc.dg/vect/vect-early-break_123-pr114297.c: New testcase.

[Bug tree-optimization/114297] [14 Regression] Yet more problems with "definition in block does not dominate use in block"

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114297

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

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

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