https://gcc.gnu.org/bugzilla/show_bug.cgi?id=125755
Bug ID: 125755
Summary: `max <min <a, nonneg>, 0>` can be simplified into
(signed)min<(unsigned)a, (unsigned)nonneg>
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 f(int a)
{
int b = 0;
int c = 100;
int t = a >= b ? a : b;
int t2 = t >= c ? c : t;
return t2;
}
```
This produces right now:
_2 = MIN_EXPR <a_3(D), 100>;
_1 = MAX_EXPR <_2, 0>;
But MAX_EXPR <_2, 0> is the same as (signed)MIN<(unsigned)_2, INT_MAX> so
the overall expression can be simplified into:
(signed)MIN_EXPR <(unsigned)a_3, 100>
This is true for all non-negative c and not just constants. 0 works because
umin with 0 argument is always 0.
This is similar to PR 122617 but not exactly.
