Missed Optimization: fold max(0, a-b) + (b-x) == 0 into select(a<0 ? b : a) == x

```llvm
define i1 @src(i64 %arg0, ptr %arg1) {
%v0 = getelementptr inbounds nuw i8, ptr %arg1, i64 72
%v1 = load i64, ptr %v0, align 8
%v2 = getelementptr inbounds nuw i8, ptr %arg1, i64 64
%v3 = load i64, ptr %v2, align 8
%v4 = sub i64 %v1, %v3
%v5 = icmp slt i64 %v1, 0
%v6 = select i1 %v5, i64 0, i64 %v4
%v7 = sub i64 %v3, %arg0
%v8 = add i64 %v7, %v6
%v9 = icmp eq i64 %v8, 0
ret i1 %v9
}

define i1 @tgt(i64 %arg0, ptr %arg1) {
%v0 = getelementptr inbounds nuw i8, ptr %arg1, i64 72
%v1 = load i64, ptr %v0, align 8
%v2 = getelementptr inbounds nuw i8, ptr %arg1, i64 64
%v3 = load i64, ptr %v2, align 8
%v4 = icmp slt i64 %v1, 0
%v5 = select i1 %v4, i64 %v3, i64 %v1
%v6 = sub i64 %v5, %arg0
%v7 = icmp eq i64 %v6, 0
ret i1 %v7
}
```

alive2: https://alive2.llvm.org/ce/z/Y2Bndb
godbolt: https://godbolt.org/z/GxavrPrP6
Pattern found in: https://github.com/dtcxzyw/llvm-opt-benchmark/blob/main/bench/cpython/optimized/bufferedio.ll