Generics: Using runtime.FuncForPC() to get method name doesn't *seem* to work

[John]

Or maybe I'm just doing something wrong.

I have a struct implementing generics where methods return the next method to execute.

For tests, I wanted to make sure the returned method was the one expected, so  I was using this:

nextStageName := runtime.FuncForPC(reflect.ValueOf(nextStage).Pointer()).Name()

(where "nextStage" is the next method) 

Here is a simple non-generic version where I get what I expect:

https://go.dev/play/p/kimGRS7xiP2

result: main.(*AStruct).NextStage-fm):

Here is probably a slightly convoluted generic version that does not:

https://go.dev/play/p/TUHUaZywki0

result: main.(*GenericStruct[...]).Stage.func1

In that version, it seem to give me the name of the method that returned the next method.

Any idea what is (or what I'm doing) wrong?

Cheers.

[Ian Lance Taylor]

I'm not sure that runtime.FuncForPC is going to work here. What it is
reporting is a function literal defined within the Stage method (that
is what the "func1" means). What is happening is that the Stage
method is returning a function literal that invokes the NextStage
method. This is due to details of how generic functions are compiled,
and may change from release to release. In general FuncForPC is a
fairly low level operation that has no way to avoid these sorts of
compilation details. I don't know if there is a way to do what you
want to do.

Ian

[John]

Thanks Ian for being generous with your time in answering this.  I'll see if I can drum up another way to make that check work. 

Cheers and have a good rest of your day.

[John]

Note for posterity:

Other methods of method equality fall in the same problem, such as detailed in:

https://stackoverflow.com/questions/9643205/how-do-i-compare-two-functions-for-pointer-equality-in-the-latest-go-weekly

Proof: https://go.dev/play/p/aeeRKfderY2