https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114025
Bug ID: 114025
Summary: Seeming missing frange based optimizations
Product: gcc
Version: 14.0
Status: UNCONFIRMED
Keywords: missed-optimization
Severity: enhancement
Priority: P3
Component: tree-optimization
Assignee: unassigned at gcc dot gnu.org
Reporter: pinskia at gcc dot gnu.org
Target Milestone: ---
Take:
```
#include <algorithm>
#include <stdexcept>
#if 1
#define AINLINE
#else
#define AINLINE [[gnu::always_inline]] inline
#endif
class TestClass
{
public:
AINLINE void SetValue(float value);
private:
float m_Value;
};
AINLINE
void TestClass::SetValue(float value)
{
if (value >= 0.0f && value <= 100.0f) {
m_Value = value;
}
else {
throw std::out_of_range("Value must be [0, 100].");
}
}
void TestFunc(TestClass& t, float value)
{
value = std::clamp(value, 30.0f, 50.0f);
// When TestClass::SetValue is inlined, the exception throwing code is not
eliminated.
// Given that at this point we can prove that 'value' lies in the range
[30.0f, 50.0f] well within the range required by the setter function, we can
rid the not taken paths of code.
t.SetValue(value);
}
```
at `-O3 -ffinite-math-only` when AINLINE is declared as nothing, TestFunc is
not always to just the clamp, but if we define it to be `inline` with
always_inline, then it is optimized to clamp. as far as I can tell the IR in
VRP1 is not able to handle correctly put the frange in for some of the SSA_NAME
and so `value >= 0.0f && value <= 100.0f` is not optimized away.
