tech-kern archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: Closing a serial device takes one second

Am 06.02.14 21:08, schrieb Dennis Ferguson:
> On 6 Feb, 2014, at 11:16 , Matthias Drochner 
> <> wrote:
>> On Thu, 6 Feb 2014 19:50:04 +0100
>> Marc Balmer <> wrote:
>>> If I find the spot, I will probably remove it.
>> It is line 784 in dev/ic/com.c. It depends on the HUPCL
>> flag unset which is done around line 1344, depending
>> on SOFTCAR.
>> The comment in com_shutdown() makes some sense, but
>> it could be done more elegantly, eg set a timer and
>> do the delay on next open if it is done immediately.
> I think the logic is broken in any case, it shouldn't
> be doing this in com_shutdown().  Notice that com_shutdown()
> isn't always executed on last close (which is when the
> documentation says the HUPCL flag is supposed to have an
> effect), it is executed on last close only when no one is
> waiting for open, e.g. waiting for the dial-in device to
> go active.  Yet the case where someone is waiting for the
> dial-in device is exactly the situation where you want to
> make sure the previous phone call is hung up, so the dial-in
> guy doesn't end up talking to the call the dial-out guy
> made.
> com_shutdown() shuts down the hardware.  You only want to
> shut down the hardware when no one has the device open and
> no one is waiting for it to be open (i.e. the condition
> under which it is called is correct), but if HUPCL is set
> you want to drop DTR on last close whether anyone is waiting
> or not.
> I hence think the delay needs to be removed from com_shutdown()
> in any case, and moved somewhere else, for this to work
> correctly.  If you just removed the delay from com_shutdown()
> and didn't move it somewhere else, however, it might be that
> no one would ever notice...

Yes, in com_shutdown() the delay can more or less safely be removed.

Which is what I plan to do.

Home | Main Index | Thread Index | Old Index