<html>
<head>
<meta content="text/html; charset=windows-1252"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<p><br>
</p>
<br>
<div class="moz-cite-prefix">On 17/07/16 00:58, Greg Woods wrote:<br>
</div>
<blockquote
cite="mid:CAF-=yYPQL+jhHZ69JX=hUQDNJ48Zs5=oUUr=F2i-=ksy81bn_g@mail.gmail.com"
type="cite">
<div dir="ltr"><br>
<div class="gmail_extra"><br>
<div class="gmail_quote">On Tue, Jul 5, 2016 at 7:07 PM, Jerry
<span dir="ltr"><<a moz-do-not-send="true"
href="mailto:mythtv@hambone.e4ward.com" target="_blank">mythtv@hambone.e4ward.com</a>></span>
wrote:<br>
<blockquote class="gmail_quote" style="margin:0px 0px 0px
0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">
<div dir="ltr">
<div class="gmail_extra">Changing the driver from
devinput to default in lirc_options.conf fixed the
issue.<br>
</div>
</div>
</blockquote>
</div>
<br>
</div>
<div class="gmail_extra">I am also having trouble getting my
remote to work well after upgrading from F23 to F24. It's the
same damned problem I had five years ago that started a long
thread on this list with F14, but it's not the same fix. That
one turned out to be an actual kernel driver bug that got
fixed. This time, most likely not.</div>
<div class="gmail_extra"><br>
</div>
<div class="gmail_extra">I've got a Streamzap remote. I applied
Jerry's fix above which got it to work, but now I'm getting
key bouncing (one press results in two actions). I am already
turning off other protocols besides lirc:</div>
<div class="gmail_extra"><br>
</div>
<div class="gmail_extra"># cat /sys/class/rc/rc0/protocols</div>
<div class="gmail_extra">other unknown rc-5 nec rc-6 jvc sony
rc-5-sz sanyo sharp mce_kbd xmp [lirc]</div>
<div class="gmail_extra"><br>
</div>
<br>
</div>
</blockquote>
This doesn't really make sense. The "fix" is really about using the
devinput driver instead of the default driver (BTW, the *default*
driver has an unfortunate name, it's a specific driver with this
name). The fix also is about how the decoding from raw IR timings
to keypress events is done.<br>
<br>
Earlier, before the "fix" Jerry used the default driver. This means
using a /dev/lirc? device and occasionally also configuring this to
use the lirc protocol. Doing so also means that the kernel decoding
is disabled and lircd decodes the raw timing info.<br>
<br>
After the "fix" Jerry is using the devinput driver. This means that
the kernel is doing the decoding, lirc is just used to distribute
the keypress events to applications (adding some functionality). In
this case the protocol should *not* be set to lirc, it should be
left untouched so the kernel can setup the internal decoding. <br>
<br>
It also means that the debug tool to be used is ir-keytable. This
reports the decoded keypress events which lircd consumes using the
devinput driver. The key question is if the events as seen by
ir-keytable is doubled. If so, it's a kernel issue.<br>
<br>
If the kernel has problems, one possible route is to use the default
driver as before. Editing lirc_options.conf to use driver=default,
device=/dev/lirc0 and dropping the streamzap config file in
/etc/lirc/lircd.conf.d should work. Mixing up these approaches
means troubles ahead...<br>
<br>
<br>
Cheers!<br>
<br>
--alec<br>
</body>
</html>