cc: [email protected] Subject: Re: Re: Re: [ast-developers] [patch] kill(1) |sigqueue()| fixes+|EAGAIN| handling etc. ... / was: Re: |sigqueue()| fixes+|EAGAIN| handling etc. ... --------
> We do have the need. The point is to queue signals in any state, even > the stopped one. The scripts we use have many worker jobs (5000+ for > small cases) which work on tasks submitted via SIGRT to it and return > to the stopped state after sending a SIGRT signal to return the result > (or better: A number defining the result directory). Once stopped they > await new instructions, again send via SIGRT. Once all tasks have been > submitted to the process we want to send a SIGCONT to kick off the > process once again to do it's stuff until the task queue has been > drained, at which the process sends itself a SIGSTOP. The parent > process receives a SIGCHLD with CHLD_STOP as notification that the > process awaits new instructions. > > Ced > OK, if there is a need for it, then I can add an option to not send SIGCONT which will remain the default. David Korn [email protected] _______________________________________________ ast-developers mailing list [email protected] http://lists.research.att.com/mailman/listinfo/ast-developers
