Re: Realtime safe signalling?
- From: Paul Davis <paul linuxaudiosystems com>
- To: Jan Hudec <bulb ucw cz>
- Cc: gtkmm-list gnome org
- Subject: Re: Realtime safe signalling?
- Date: Mon, 21 Feb 2005 10:04:30 -0500
>The writer of the manual could not guarantee how will write(2) behave,
>since the code is system-independent. Linux does protect pipe with
>a semaphore. On single-CPU, it can never be waited for unless the buffer
the FIFO semaphore is not the central issue, particular if the FIFO is
being used for single reader/single writer communication.
the problem is if the FIFO is created in a filesystem that is updating
metadata (such as modification time or a journal), where access to the
FIFO will require taking the filesystem lock too.
it is for this reason that all packaged version of JACK
(jackit.sf.net) create a tmpfs filesystem specifically as a place to
create the FIFOs used by jackd when communicating with clients. and
JACK is about as hard-RT as linux apps get.
--p
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]