:: Re: [DNG] Doing away with multi-thr…
トップ ページ
このメッセージを削除
このメッセージに返信
著者: Edward Bartolo
日付:  
To: Rainer Weikusat
CC: dng
題目: Re: [DNG] Doing away with multi-threading in my project (netman)
A periodic algorithm is now implemented in the frontend to search for
backend and any spawned children for ownership by root. This most
probably was what was causing waitpid and wait to fail to reap
zombies. If the algorithm find a root process parented by the
frontend, it will disable 'connect' and 'disconnect' until the process
dies. This prevents against having multiple threads initiated by
impatient users.

The next step is semi/automatic wifi connection.


Edward

On 04/09/2015, Rainer Weikusat <rainerweikusat@???> wrote:
> Edward Bartolo <edbarx@???> writes:
>
>> Hi,
>>
>> For the last days I have been struggling to reap zombies adopted by
>> the frontend with only failures.
>
> The frontend doesn't adopt zombies. Only init ever does that.
>
>> I tried several methods using wait(-1), waidpid and trapping the
>> signal when a child dies without ever succeeding to reap zombies. This
>> means, I will stop and will leave the issue open to anyone who can
>> find a working solution.
>
> The 'zombie reaping' needs to happen in the parent process. This means
> the backend has to care for any processes it starts via fork. And the
> frontend has to care for the backend. You can get a tree view of
> processes via ps f, eg,
>
> ps fx
>
> to show all processed running under you current uid.
>
> [...]
>
>
>> initialization
>>
>> sa.sa_handler := @handle_sigchld;
>
> Insofar the documentation goes, you should be able to replace this with
>
> sa.sa_handler := SIG_IGN;
>
> this means now direct backend processes will ever be zombified. As the
> backend is written in C, a single
>
> signal(SIGCHLD, SIG_IGN);
>
> will take care of all processes spawned by the backend.
> _______________________________________________
> Dng mailing list
> Dng@???
> https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
>