[Bug tree-optimization/125385] New: Expand ((float)a) CMP ((float)b) to use the ranges rather than the just type

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

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=125385

            Bug ID: 125385
           Summary: Expand ((float)a) CMP ((float)b) to use the ranges
                    rather than the just type
           Product: gcc
           Version: 16.0
            Status: UNCONFIRMED
          Keywords: easyhack, missed-optimization
          Severity: normal
          Priority: P3
         Component: tree-optimization
          Assignee: unassigned at gcc dot gnu.org
          Reporter: pinskia at gcc dot gnu.org
  Target Milestone: ---

Take:
```
int f0(int a, int b)
{
  a &=0xff;
  b &=0xff;
  return (float) a < (float) b;
}


int f1(int a, int b)
{
  unsigned char aa = a;
  unsigned char bb = b;
  return (float) aa < (float) bb;
}
```

Both of these should optimized to the same but currently does not since we only
look at the precission of the type of int rather than the range of a to see we
could fit it into a float.