[erlang-questions] gen_statem 20.3 init/1 - {next_event, internal, initial_event}

[Raimo Niskanen]

The fix is part of the newly released emergency patch package OTP-20.3.2.
/ Raimo


On Thu, Mar 15, 2018 at 03:17:14PM +0100, Peter Morgan wrote:
> Thanks very much for the diff!
> Regards
> Peter 
> 
> > On 15 Mar 2018, at 14:51, Raimo Niskanen <raimo+erlang-questions@REDACTED> wrote:
> > 
> > I am sorry i fubmled when optimizing gen_statem.  That typo is mine,
> > besides the erroneous actions list handling from init/1.
> > 
> > This diff should fix the problems:
> > 
> > diff --git a/lib/stdlib/src/gen_statem.erl b/lib/stdlib/src/gen_statem.erl
> > index b1d99e2..eb0d6bd 100644
> > --- a/lib/stdlib/src/gen_statem.erl
> > +++ b/lib/stdlib/src/gen_statem.erl
> > @@ -677,9 +677,9 @@ enter(Module, Opts, State, Data, Server, Actions,
> > Parent) ->
> >     NewDebug = ?sys_debug(Debug, {Name,State}, {enter,Event,State}),
> >     case call_callback_mode(S) of
> >        #state{} = NewS ->
> > -           loop_event_actions(
> > +           loop_event_actions_list(
> >              Parent, NewDebug, NewS,
> > -             Events, Event, State, Data, #trans_opts{},
> > +             Events, Event, State, Data, false,
> >               NewActions, CallEnter);
> >        [Class,Reason,Stacktrace] ->
> >            terminate(
> > @@ -1295,7 +1295,7 @@ parse_actions_next_event(
> >                 next_events_r = [{Type,Content}|NextEventsR]});
> >         _ ->
> >             [error,
> > -             {bad_action_from_state_function,{next_events,Type,Content}},
> > +             {bad_action_from_state_function,{next_event,Type,Content}},
> >              ?STACKTRACE(),
> >              ?not_sys_debug]
> >     end;
> > @@ -1312,7 +1312,7 @@ parse_actions_next_event(
> >                 next_events_r = [{Type,Content}|NextEventsR]});
> >         _ ->
> >             [error,
> > -             {bad_action_from_state_function,{next_events,Type,Content}},
> > +             {bad_action_from_state_function,{next_event,Type,Content}},
> >              ?STACKTRACE(),
> >              Debug]
> >     end.
> > 
> > 
> > 
> > That, running it through our daily builds, and apparently writing a
> > writing a missing test case. ;-)
> > 
> > I am at the San Francisco Code Beam event right now so a patch will have to
> > wait to next week..
> > 
> > Sorry about this!
> > / Raimo Niskanen
> > 
> > 
> > 
> >> On Thu, Mar 15, 2018 at 12:32:04PM +0000, Guilherme Andrade wrote:
> >> Crash log mentions 'next_events' (rather than 'next_event'.) Perhaps you
> >> have a typo somewhere? But if that's the case, then it's weird that 20.2.x
> >> doesn't complaint about it as well.
> >> 
> >> On 15 March 2018 at 11:23, Peter Morgan <peter.james.morgan@REDACTED>
> >> wrote:
> >> 
> >>> Hello -
> >>> 
> >>> I am sure I am doing something daft. I picked up 20.3 this morning and
> >>> have some gen_statem code that previously worked in 20.2.4:
> >>> 
> >>> -module(ex1).
> >>> 
> >>> -export([callback_mode/0]).
> >>> -export([handle_event/4]).
> >>> -export([init/1]).
> >>> -export([start/0]).
> >>> 
> >>> start() ->
> >>>    gen_statem:start({local, ?MODULE}, ?MODULE, [], []).
> >>> 
> >>> init([]) ->
> >>>    {ok,
> >>>     initial_state,
> >>>     initial_data,
> >>>     {next_event, internal, initial_event}}.
> >>> 
> >>> callback_mode() ->
> >>>    handle_event_function.
> >>> 
> >>> handle_event(Type, Event, State, Data) ->
> >>>    error_logger:info_report([{type, Type},
> >>>                              {event, Event},
> >>>                              {state, State},
> >>>                              {data, Data}]),
> >>>    keep_state_and_data.
> >>> 
> >>> 
> >>> 
> >>> 
> >>> brew switch erlang 20.2.4
> >>> 
> >>> Erlang/OTP 20 [erts-9.2.1] [source] [64-bit] [smp:4:4] [ds:4:4:10]
> >>> [async-threads:10] [hipe] [kernel-poll:false] [dtrace]
> >>> 
> >>> Eshell V9.2.1  (abort with ^G)
> >>> 1> ex1:start().
> >>> {ok,<0.62.0>}
> >>> 
> >>> =INFO REPORT==== 15-Mar-2018::12:13:29 ===
> >>>    type: internal
> >>>    event: initial_event
> >>>    state: initial_state
> >>>    data: initial_data
> >>> 
> >>> 
> >>> 
> >>> brew switch erlang 20.3
> >>> Erlang/OTP 20 [erts-9.3] [source] [64-bit] [smp:4:4] [ds:4:4:10]
> >>> [async-threads:10] [hipe] [kernel-poll:false] [dtrace]
> >>> 
> >>> Eshell V9.3  (abort with ^G)
> >>> 1> c(ex1).
> >>> {ok,ex1}
> >>> 2> ex1:start().
> >>> {ok,<0.67.0>}
> >>> 3>
> >>> =ERROR REPORT==== 15-Mar-2018::12:15:02 ===
> >>> ** State machine ex1 terminating
> >>> ** Last event = {internal,init_state}
> >>> ** When server state  = {initial_state,initial_data}
> >>> ** Reason for termination = error:{bad_action_from_state_function,
> >>>                                      {next_events,internal,
> >>> initial_event}}
> >>> ** Callback mode = handle_event_function
> >>> ** Stacktrace =
> >>> **  [{gen_statem,parse_actions_next_event,7,
> >>>                 [{file,"gen_statem.erl"},{line,1299}]},
> >>>     {gen_statem,loop_event_actions_list,10,
> >>>                 [{file,"gen_statem.erl"},{line,1194}]},
> >>>     {proc_lib,init_p_do_apply,3,[{file,"proc_lib.erl"},{line,247}]}]
> >>> 
> >>> 
> >>> The docs seem to indicate that next_event remains valid on init/1 -
> >>> http://erlang.org/doc/man/gen_statem.html#type-init_result.
> >>> 
> >>> Thanks
> >>> Peter.
> >>> 
> >>> 
> >>> _______________________________________________
> >>> erlang-questions mailing list
> >>> erlang-questions@REDACTED
> >>> http://erlang.org/mailman/listinfo/erlang-questions
> >>> 
> >>> 
> >> 
> >> 
> >> -- 
> >> Guilherme
> > 
> >> _______________________________________________
> >> erlang-questions mailing list
> >> erlang-questions@REDACTED
> >> http://erlang.org/mailman/listinfo/erlang-questions
> > 
> > 
> > -- 
> > 
> > / Raimo Niskanen, Erlang/OTP, Ericsson AB
> > _______________________________________________
> > erlang-questions mailing list
> > erlang-questions@REDACTED
> > http://erlang.org/mailman/listinfo/erlang-questions

-- 

/ Raimo Niskanen, Erlang/OTP, Ericsson AB