Update on "[Inductor][CPP] Add Min/Max with VecMask"

**Summary**
Fix issue: #126824 which is missing the support of `min/max` with `VecMask`.

**TestPlan**
```
python test/inductor/test_torchinductor_opinfo.py -k test_comprehensive_clamp_max_cpu_bool
python test/inductor/test_torchinductor_opinfo.py -k test_comprehensive_clamp_min_cpu_bool
```

cc jgong5 mingfeima XiaobingSuper sanchitintel ashokei jingxu10 voznesenskym penguinwu EikanWang Guobing-Chen zhuhaozhe blzheng wenzhe-nrv jiayisunx peterbell10 ipiszy yf225 chenyang78 kadeng muchulee8 ColinPeppler amjames desertfire chauhang

[ghstack-poisoned]

aten/src/ATen/cpu/vec/vec_mask.h

@@ -145,18 +145,6 @@ class VecMask {
     return result;
   }
 
-  static VecMask<T, N> minimum(
-    const VecMask<T, N>& a,
-    const VecMask<T, N>& b) {
-    return VecMask<T, N>::blendv(b, a, a < b);
-  }
-
-  static VecMask<T, N> maximum(
-    const VecMask<T, N>& a,
-    const VecMask<T, N>& b) {
-    return VecMask<T, N>::blendv(b, a, a > b);
-  }
-
   void store(bool* b, int count = size()) {
     constexpr int L = (VectorizedN<T, N>::size() + Vectorized<bool>::size() - 1)/ Vectorized<bool>::size();
     auto res = this->to<bool, L>();

torch/_inductor/codegen/cpp.py

@@ -1318,7 +1318,7 @@ def minimum(a, b):
         if a.dtype == torch.bool:
             assert b.dtype == torch.bool
             a_cast, b_cast = unify_mask_base_type(V.kernel.compute, (a, b))
-            return f"decltype({a_cast})::minimum({a_cast}, {b_cast})"
+            return f"{a_cast} & {b_cast}"
         else:
             return f"at::vec::minimum({a}, {b})"
 
@@ -1327,7 +1327,7 @@ def maximum(a, b):
         if a.dtype == torch.bool:
             assert b.dtype == torch.bool
             a_cast, b_cast = unify_mask_base_type(V.kernel.compute, (a, b))
-            return f"decltype({a_cast})::maximum({a_cast}, {b_cast})"
+            return f"{a_cast} | {b_cast}"
         else:
             return f"at::vec::maximum({a}, {b})"