mixing signed and unsigned

[T]

Given:

typedef unsigned short u16;
typedef signed int i32;

u16 u0 = 0xFFFF;
i32 i1 = u0;

u16 u1 = 0;
i32 i2 = u1 - (u16)1;

printf("p1: %d, %d \n", i1, i2 );
-------------------------------------------------------

Why do i1 & i2 have different results?

Isn't the rhs of i2 supposed to be done entirely in u16?

or put another way, why isn't the above i2 computed like this
i2 = u16( u1 - (u16)1)

 

[Alf P. Steinbach]

* T:

Given:

typedef unsigned short u16;
typedef signed int i32;

u16 u0 = 0xFFFF;
i32 i1 = u0;

u16 u1 = 0;
i32 i2 = u1 - (u16)1;

printf("p1: %d, %d \n", i1, i2 );
-------------------------------------------------------

Why do i1 & i2 have different results?

Isn't the rhs of i2 supposed to be done entirely in u16?

No. It's done in 'int'. Integer promotions.

 

[T]

then why isn't u0 promoted in the same manner?

 

[Alf P. Steinbach]

* T:

then why isn't u0 promoted in the same manner?

I'd love to answer this, but I have no idea what you're talking about
since you didn't quote anything.