On Sat, Jul 30, 2022 at 7:26 AM Nobishii N <nobishi...@gmail.com> wrote: > > I have read Go1.19(tip) memory model document(https://tip.golang.org/ref/mem) > and have following question. (I think my question is valid for the current > model(https://go.dev/ref/mem), though.) > > Here is my translation of so-called "Message Passing" litmus test to Go > program. > > https://go.dev/play/p/hPn7yrIN4Sf > > Does this program can panic according to the memory model? > I think it can panic according to the model, but I'm not confident because it > is subtle problem for me. > > My reasoning is as follows. > > First of all, this program has data race. Behavior for data races are > documented in the section "Implementation Restrictions for Programs > Containing Data Races". According to this section, > > - read y := y (L18) can observe both y = 1(L13) and y = 0(initialization at > L5) > - read x := x (L19) can observe both x = 1(L12) and x = 0(initialization at > L5) > > so the result "x == 0 && y == 1" can actually happen, causing this program to > panic. > > Is my reasoning correct?
I agree that this program can panic. Ian -- 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+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/golang-nuts/CAOyqgcUmDBk6t43q%3DKebdFJ-yYZjnfVvpFse71pzB_HSv%2Bw8dw%40mail.gmail.com.
