[Bug tree-optimization/121881] New: cselim does not handle some stuff sometimes

[pinskia at gcc dot gnu.org via Gcc-bugs]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=121881

            Bug ID: 121881
           Summary: cselim does not handle some stuff sometimes
           Product: gcc
           Version: 16.0
            Status: UNCONFIRMED
          Keywords: missed-optimization
          Severity: normal
          Priority: P3
         Component: tree-optimization
          Assignee: pinskia at gcc dot gnu.org
          Reporter: pinskia at gcc dot gnu.org
  Target Milestone: ---

Take:
```
struct s1
{
  int t;
};

struct s1 f(int a, int b)
{
  struct s1 t;
  struct s1 c;
  int *c1 = &t.t;
  if (a)
    t.t = b;
  else
    *c1 = b;
  return t;
}
```

Compile at `-O2 -fno-tree-sra` and GCC produces:
```
  <bb 2> [local count: 1073741824]:
  if (a_2(D) != 0)
    goto <bb 3>; [50.00%]
  else
    goto <bb 4>; [50.00%]

  <bb 3> [local count: 536870912]:
  # .MEM_6 = VDEF <.MEM_3(D)>
  t.t = b_4(D);
  goto <bb 5>; [100.00%]

  <bb 4> [local count: 536870912]:
  # .MEM_5 = VDEF <.MEM_3(D)>
  MEM[(int *)&t] = b_4(D);

  <bb 5> [local count: 1073741824]:
  # .MEM_1 = PHI <.MEM_6(3), .MEM_5(4)>
  # .MEM_7 = VDEF <.MEM_1>
```

BUT `t.t` and `MEM[(int *)&t]` point to the same location just a different
representation of the stores.

I Noticed this in the IR from PR 119482 (ladybird, this is with SRA turned on
even):
```
  <bb 13> [local count: 66534413]:
  # SR.3993_106 = PHI <9222809086901354496(11), SR.3993_63(12)>
  iftmp.1316.D.157125.m_value.encoded = SR.3993_106;
  goto <bb 15>; [100.00%]

  <bb 14> [local count: 66534413]:
  MEM[(struct Value *)&iftmp.1316] = 9222809086901354496;

  <bb 15> [local count: 133068825]:
```

Both of these are the same stores.