[Bug tree-optimization/110768] New: [14 Regression] Dead Code Elimination Regression since r14-2623-gc11a3aedec2

[Bug tree-optimization/110768] New: [14 Regression] Dead Code Elimination Regression since r14-2623-gc11a3aedec2

[theodort at inf dot ethz.ch]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110768

            Bug ID: 110768
           Summary: [14 Regression] Dead Code Elimination Regression since
                    r14-2623-gc11a3aedec2
           Product: gcc
           Version: 14.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: tree-optimization
          Assignee: unassigned at gcc dot gnu.org
          Reporter: theodort at inf dot ethz.ch
  Target Milestone: ---

https://godbolt.org/z/GsTMz1G9c

Given the following code:

void foo(void);
static int a, b;
int main() {
    {
        short c = 45127;
        char d;
        b = 0;
        for (; b <= 3; b++) {
            if (b) continue;
            d = 0;
            for (; d <= 3; d++) {
                if (!(((c) >= -20409) && ((c) <= 1))) {
                    __builtin_unreachable();
                }
                if (~(0 == a) & 1) return b;
                c = 0;
                for (; c <= 0; c++) a = 3;
            }
        }
        foo();
    }
}

gcc-trunk -Os does not eliminate the call to foo:

main:
        xorl    %r9d, %r9d
        movl    a(%rip), %eax
        xorl    %edx, %edx
        xorl    %esi, %esi
        movl    %r9d, b(%rip)
        xorl    %ecx, %ecx
.L2:
        cmpl    $4, %edx
        je      .L31
        testl   %edx, %edx
        jne     .L3
        testl   %eax, %eax
        je      .L4
.L24:
        testb   %sil, %sil
        je      .L5
        xorl    %r8d, %r8d
        movl    %r8d, b(%rip)
.L5:
        testb   %cl, %cl
        je      .L27
        movl    %eax, a(%rip)
        jmp     .L27
.L4:
        movl    $3, %eax
        movb    $1, %cl
        jmp     .L24
.L3:
        incl    %edx
        movb    $1, %sil
        jmp     .L2
.L31:
        pushq   %rdi
        testb   %sil, %sil
        je      .L9
        movl    $4, b(%rip)
.L9:
        testb   %cl, %cl
        je      .L10
        movl    %eax, a(%rip)
.L10:
        call    foo
        xorl    %eax, %eax
        popq    %rdx
        ret
.L27:
        xorl    %eax, %eax
        ret

gcc-13.1.0 -Os eliminates the call to foo:

main:
        xorl    %eax, %eax
        cmpl    $0, a(%rip)
        movl    %eax, b(%rip)
        jne     .L2
        movl    $3, a(%rip)
.L2:
        xorl    %eax, %eax
        ret

Bisects to r14-2623-gc11a3aedec2

[Bug tree-optimization/110768] [14 Regression] Dead Code Elimination Regression since r14-2623-gc11a3aedec2

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110768

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |missed-optimization
   Target Milestone|---                         |14.0

[Bug tree-optimization/110768] [14 Regression] Dead Code Elimination Regression since r14-2623-gc11a3aedec2

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110768

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

  3) optimize_size logic is now different.  Originally we started duplicating
     iff the first conditional was known to be true by ranger query, but then
     we used same limits as for -O2.

[Bug tree-optimization/110768] [14 Regression] Dead Code Elimination Regression since r14-2623-gc11a3aedec2

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110768

--- Comment #2 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Looks to be fixed now.

[Bug tree-optimization/110768] [14 Regression] Dead Code Elimination Regression since r14-2623-gc11a3aedec2

[jamborm at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110768

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jamborm at gcc dot gnu.org
           Keywords|needs-bisection             |

--- Comment #3 from Martin Jambor <jamborm at gcc dot gnu.org> ---
This has been fixed with r14-5109-ga291237b628f41 (Andrew MacLeod: Remove
simple ranges from trailing zero bitmasks)

[Bug tree-optimization/110768] [14 Regression] Dead Code Elimination Regression since r14-2623-gc11a3aedec2

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110768

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
     Ever confirmed|0                           |1
           Assignee|unassigned at gcc dot gnu.org      |pinskia at gcc dot gnu.org
             Status|UNCONFIRMED                 |ASSIGNED
   Last reconfirmed|                            |2024-01-11

--- Comment #4 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Will commit the testcase and close it as fixed.

The range improvements allows loop ch to do a better job ...

[Bug tree-optimization/110768] [14 Regression] Dead Code Elimination Regression since r14-2623-gc11a3aedec2

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110768

--- Comment #5 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Note the main difference after this patch:
Global Exported: c_11 = [irange] short int [-20409, -20409][1, 1] MASK 0xb046
VALUE 0xb047


vs before:
Global Exported: c_11 = [irange] short int [-INF, +INF] NONZERO 0xb047


That is we figure out c_11 will only either be -20409 [(short)45127] or 1 now.

That allows to get rid of if statement that leads to unreachable.
Which then allows unrolling to happen as the estimate for that loop is less.

Note the following testcase (which is fixed also on the trunk) was failing in
GCC 13:
```
void foo(void);
static int a, b;
int main() {
    {
        short c = 45127;
        int d;
        b = 0;
        for (; b <= 3; b++) {
            if (b) continue;
            d = 0;
            for (; d <= 127; d++) {
                if (!(((c) >= -20409) && ((c) <= 1))) {
                    __builtin_unreachable();
                }
                if (~(0 == a) & 1) return b;
                c = 0;
                for (; c <= 0; c++) a = 3;
            }
        }
        foo();
    }
}

```

The difference between this one and the original testcase is the # of
iterations of the loop. The same thing applies here; though I don't know what
caused the regression between GCC 12 and 13 for this one.

[Bug tree-optimization/110768] [14 Regression] Dead Code Elimination Regression since r14-2623-gc11a3aedec2

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110768

--- Comment #6 from GCC Commits <cvs-commit at gcc dot gnu.org> ---
The trunk branch has been updated by Andrew Pinski <pinskia@gcc.gnu.org>:

https://gcc.gnu.org/g:34a827039fabcf24ce78da25984a1cc8be7ca2c3

commit r14-7221-g34a827039fabcf24ce78da25984a1cc8be7ca2c3
Author: Andrew Pinski <quic_apinski@quicinc.com>
Date:   Fri Jan 12 20:24:34 2024 -0800

    Add a few testcases for fix missed optimization regressions

    Adds a few new testcases for some missed optimization regressions.
    The analysis on how each should be optimized is in the testcases
    themselves (and in the bug report).

    Committed as obvious after running the testsuite to make sure they pass.

            PR tree-optimization/107823
            PR tree-optimization/110768
            PR tree-optimization/110941
            PR tree-optimization/110450
            PR tree-optimization/110841

    gcc/testsuite/ChangeLog:

            * gcc.dg/tree-ssa/ssa-thread-22.c: New test.
            * gcc.dg/tree-ssa/vrp-loop-1.c: New test.
            * gcc.dg/tree-ssa/vrp-loop-2.c: New test.
            * gcc.dg/tree-ssa/vrp-unreachable-1.c: New test.
            * gcc.dg/tree-ssa/vrp-unreachable-2.c: New test.

    Signed-off-by: Andrew Pinski <quic_apinski@quicinc.com>

[Bug tree-optimization/110768] [14 Regression] Dead Code Elimination Regression since r14-2623-gc11a3aedec2

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110768

--- Comment #7 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Fixed.

[Bug tree-optimization/110768] [14 Regression] Dead Code Elimination Regression since r14-2623-gc11a3aedec2

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110768

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

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

--- Comment #8 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Fixed.