Behavior to copy the value of pointer inside another pointer

[Olivier Szika]

Hi all,

I am surprised by the behavior of "&(*var)".

https://play.golang.org/p/CK522lRkddd

It is expected?

[burak serdar]

On Wed, Aug 4, 2021 at 9:18 AM Olivier Szika <szika....@gmail.com> wrote:

Hi all,

I am surprised by the behavior of "&(*var)".

According to the language spec, pointer indirection is addressable:

https://golang.org/ref/spec#Address_operators

Thus, taking the address of an indirection yields the original pointer, not a pointer to a copy of its contents.

 

https://play.golang.org/p/CK522lRkddd

It is expected?

--
You received this message because you are subscribed to the Google Groups "golang-nuts" group.
To unsubscribe from this group and stop receiving emails from it, send an email to golang-nuts...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/golang-nuts/5f265699-ed67-4798-92d0-88832a7444d1n%40googlegroups.com.

[Axel Wagner]

Yes. From the spec:

For an operand x of pointer type *T, the pointer indirection *x denotes the variable of type T pointed to by x.

Emphasis mine. "The variable", means it really denotes the storage location that `x` points to. Not the value or anything like that. Further, then:

For an operand x of type T, the address operation &x generates a pointer of type *T to x. The operand must be addressable, that is, either a variable, pointer indirection, or slice indexing operation; or a field selector of an addressable struct operand; or an array indexing operation of an addressable array. As an exception to the addressability requirement, x may also be a (possibly parenthesized) composite literal.

 

Emphasis mine, again. This tells you that the behavior of `&` with composite literals to allocate a new variable is an exception - in general, `&` just gives you the address of whatever expression you apply it to. In this case, specifically, a pointer-indirection denoting, as we've seen above, the variable that x points to.

So, it is completely expected that `&(*p) == p`.

--

[Olivier Szika]

Oh Yes,

Thanks!