<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7655.5">
<TITLE>Re: I2C RW failure, with experimental 80e driver...</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/plain format -->

<P><FONT SIZE=2>Hi Devin -<BR>
<BR>
&nbsp;&nbsp; I will look into the I2C comms then.&nbsp; I also have access to the source, so I'm not totally flying blind (however, it is a BIG code base!)<BR>
As far as playback goes, I'm using a custom player that interfaces with the V4L frontend file device.&nbsp; No Mythtv frontend.&nbsp; Although, I wish there was a vanilla mythtv app<BR>
that just allows me to tune to a frequency and play out to the monitor/HDMI/etc<BR>
<BR>
&nbsp;&nbsp; I understand you're under NDA, but therein creates a problem:&nbsp; you have access to the specs, but no time.&nbsp; I have time, but no specs.&nbsp; If anything the code should probably go into<BR>
v4l_experimental or something.&nbsp; Is there anyone in your contacts I can speak to about getting the specs (under my own NDA of course).&nbsp; Plus I have access to a Tektronix QAM/VSB generator, but I don't know for how long.<BR>
<BR>
&nbsp;&nbsp; Cheers, Charles.<BR>
<BR>
Date: Wed, 11 May 2011 09:35:08 -0400<BR>
From: Devin Heitmueller &lt;dheitmueller@kernellabs.com&gt;<BR>
Subject: Re: [mythtv-users] I2C RW failure, with experimental 80e<BR>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; driver...<BR>
To: Discussion about MythTV &lt;mythtv-users@mythtv.org&gt;<BR>
Message-ID: &lt;BANLkTim=k4970MYnV7MDm3xf7_RXHnZfNw@mail.gmail.com&gt;<BR>
Content-Type: text/plain; charset=ISO-8859-1<BR>
<BR>
On Wed, May 11, 2011 at 9:14 AM, Martin, Charles &lt;cmart@dolby.com&gt; wrote:<BR>
&gt; Hi Devin -<BR>
&gt;<BR>
&gt; ? Thanks for the fast response. ?I entirely believe you on being busy and all.<BR>
&gt; I will admit it's a case of varying mileage when I try experimental code against an older kernel, on a STB to boot!<BR>
&gt; I will enable all the I2C options and see if that will fix it.<BR>
&gt;<BR>
&gt; ? I don't recall any message of any firmwares loading up with this particular device (nothing called from request_firmware() ) in the USB stack.<BR>
&gt; I did see one quirky thing where the frontend id was assigned some garbage value. ?I put a memset statement in the drx39xyj_attach() routine, but<BR>
&gt; I'm not sure if that's the correct thing to do (because the state-&gt;fe field seems to be ptr too).<BR>
&gt;<BR>
&gt; ? Actually, I was trying to test ClearQAM reception when I tried it, so that's good to know. ?However, if you need help getting this up and running,<BR>
&gt; I could probably help, if you can give me access to chip datasheets and so forth, or give me pointers on how to proceed. ?I'll code and test, and you can<BR>
&gt; tell me what to do essentially ;-)<BR>
&gt;<BR>
&gt; ? I can be reached at wraith_dart-at-yahoo-dot-com as well.<BR>
&gt;<BR>
&gt; Cheers, C.<BR>
<BR>
Hi Charles,<BR>
<BR>
You're right - I forgot that the driver in question currently still<BR>
has the firmware embedded,&nbsp; It will need to be moved out to a separate<BR>
binary in order to get upstream, but I didn't get that far in the<BR>
driver port.&nbsp; Again, my bad for just not having poked at the code in<BR>
almost a year.<BR>
<BR>
If I can make a suggestion:&nbsp; you may find it useful to start your<BR>
process on a current Ubuntu or Fedora x86 box.&nbsp; That will allow you to<BR>
learn how the various applications and drivers work, after which point<BR>
you can then move to the embedded platform.&nbsp; It seems that you've<BR>
combined learning about how LinuxTV and MythTV work at the same time<BR>
as trying to make it work on an obscure platform, so it would be extra<BR>
difficult to understand whether your issues are with the platform<BR>
specifically or just your general lack of knowledge as to how the<BR>
devices and applications are supposed to work.<BR>
<BR>
I cannot make available any datasheets as everything is under NDA.<BR>
That said though, this should be pretty straightforward to debug if<BR>
you have any kernel driver experience (the problem appears to be<BR>
related to general i2c communications).<BR>
<BR>
Devin<BR>
<BR>
--<BR>
Devin J. Heitmueller - Kernel Labs<BR>
<A HREF="http://www.kernellabs.com">http://www.kernellabs.com</A><BR>
</FONT>
</P>

</BODY>
</HTML>