----- Mail original -----
> De: "daniel smith" <daniel.sm...@oracle.com>
> À: "Remi Forax" <fo...@univ-mlv.fr>
> Cc: "Brian Goetz" <brian.go...@oracle.com>, "amber-spec-experts"
> <amber-spec-experts@openjdk.java.net>
> Envoyé: Mercredi 11 Avril 2018 00:20:01
> Objet: Re: Switch expressions -- gathering the threads
>> On Apr 10, 2018, at 2:30 PM, Remi Forax <fo...@univ-mlv.fr> wrote:
>>
>> I'm not sure this difference is important.
>>
>> What about the example below, multiple labels or a fallthrough ?
>> switch(x) {
>> case 0:
>> ;
>> case 1:
>> }
>
> My request is to call this an example of fallthrough.
>
> I think you're trying to make a point that some forms of switches with
> fallthrough behave the same as switches with multiple labels. Sure, that's
> fine. I still think it's helpful to talk about the two cases separately, as
> distinct features, because the practical use cases are very different.
I think is see all forms as being fallthrough and what you call a multiple
labels form as the result after a peephole optimization, i.e if there is no
instruction between the two cases, then the compiler will make them share the
same label.
>
> —Dan
Rémi
