[vlc-devel] [PATCH 2/2] chromecast: don't close the client connection

[Rémi Denis-Courmont]

Le tiistaina 27. lokakuuta 2020, 19.15.44 EET Alexandre Janniaux a écrit :
> Hi,
> 
> On Tue, Oct 27, 2020 at 07:10:03PM +0200, Rémi Denis-Courmont wrote:
> > Le tiistaina 27. lokakuuta 2020, 18.28.00 EET Thomas Guillem a écrit :
> > > Fixes #25151
> > > Fixes #25192
> > > ---
> > > 
> > >  modules/stream_out/chromecast/cast.cpp | 7 +++++++
> > >  1 file changed, 7 insertions(+)
> > > 
> > > diff --git a/modules/stream_out/chromecast/cast.cpp
> > > b/modules/stream_out/chromecast/cast.cpp index 08defbd5a1b..12e175bc68a
> > > 100644
> > > --- a/modules/stream_out/chromecast/cast.cpp
> > > +++ b/modules/stream_out/chromecast/cast.cpp
> > > @@ -466,6 +466,13 @@ int sout_access_out_sys_t::url_cb(httpd_client_t
> > > *cl,
> > > httpd_message_t *answer, * should also serve data used by the old client
> > > to
> > > the new one. */ restoreCopy();
> > > 
> > >          m_client = cl;
> > > 
> > > +
> > > +        /* Chromecast clients buffer lot of data (burst mode) and can
> > > wait
> > > more
> > > +         * than 10 seconds between 2 consecutive reads. Tell httpd
> > > that the
> > > +         * timeout of this client should not be handled (infinite
> > > timeout) to
> > > +         * avoid closing the client connection after 10
> > > seconds without any
> > > +         * read. */
> > > +        httpd_ClientSetTimeout(m_client, 0);
> > 
> > In sending state, the timeout will fire if the client cannot keep up with
> > the data. That being the case, the TCP connection timeout will fire
> > anyway after some tens of seconds. You can't do infinite timeout.
> 
> That is surely incomplete to solve the underlying issue, as
> noted by François originally.

First, I have yet to see an explaination of what the underlying issue is, from 
François or anyone else.

And then, in receiving state, activity time-out is necessary to remove dead or 
inactive client and not clog the server. In sending state, the HTTPd activity 
time-out is just there to expedite client removal. The TCP time-out would 
eventually kill the connection anyway. Either way, disabling the time-out 
looks to me like adding a bug and fixing nothing.

-- 
Реми Дёни-Курмон
http://www.remlab.net/