[mythtv-users] 1212 and channel changing

[Jarod Wilson]

On Aug 23, 2009, at 2:28 PM, David Engel wrote:

> On Sun, Aug 23, 2009 at 12:32:26AM -0400, Jarod Wilson wrote:
>> On Saturday 22 August 2009 15:58:47 David Engel wrote:
>>> Second problem is lirc_zilog doesn't seem to detect the hdpvr i2c
>>> adapter.  I've added printk's to ir_probe to see what ids are  
>>> getting
>>> checked, but AFAICT, it is never getting called.
>>
>> There are some issues... The current git head lirc_zilog is for  
>> 2.6.31,
>> which includes a bunch of i2c changes. However, even with your patch,
>> I think there's still some magic missing in lirc_zilog to get it to
>> register properly to bind to lirc_zilog on 2.6.31... To work on on
>> pre-2.6.31, I think lirc_zilog needs the i2c changes reverted to
>> function with earlier kernels, I think. This patch I believe should
>> do the trick, testing it shortly...
>>
>> http://wilsonet.com/jarod/junk/lirc_zilog-revert-2.6.31-i2c-changes.patch
>
> Good catch.  I was using 2.6.30 and this patch fixed the registration
> problem such that /dev/lirc0 is created and lircd will start.

Yeah, with 2.6.29.6 and the hdpvr on my end now:

lirc_zilog: chip found with RX and TX
lirc_dev: lirc_register_driver: sample_rate: 0
i2c ir driver 0-0070: firmware: requesting haup-ir-blaster.bin
lirc_zilog: firmware of size 302355 loaded
lirc_zilog: 743 codesets loaded
lirc_zilog: Zilog/Hauppauge IR blaster: firmware version 2.1.0

> FYI,
> trying to load lirc_zilog when the firmware file doesn't exist and
> trying to unload lirc_zilog both hang.

Hrm, someone should fix that... Adding to my TODO list...

> I've never tried the blaster capabilities of lirc before, so I'm
> loosely following the instructions found at
> http://www.blushingpenguin.com/mark/blog/?p=24 .  When I tried to run
> "irsend SEND_ONCE blaster 0_1_KEY_POWER" it fails and lircd logs the
> following:
>
> Aug 23 13:02:56 opus lircd-0.8.3[3999]: accepted new client on /dev/ 
> lircd
> Aug 23 13:02:58 opus lircd-0.8.3[3999]: write failed
> Aug 23 13:02:58 opus lircd-0.8.3[3999]: Bad address
> Aug 23 13:02:58 opus lircd-0.8.3[3999]: error processing command:  
> SEND_ONCE blaster 0_1_KEY_POWER
> Aug 23 13:02:58 opus lircd-0.8.3[3999]: transmission failed
> Aug 23 13:02:58 opus lircd-0.8.3[3999]: removed client

You're further along than I am, I've only loaded the driver and  
confirmed it binds thus far. Will have to do some poking to see if  
anything works tonight.

Note that I have some suspicions about compatibility between a lirc  
0.8.3 userspace and more recent kernel drivers. I know lirc 0.8.5 with  
older kernel drivers fails spectacularly, it seems likely lirc 0.8.3  
with newer kernel drivers is also subject to fail.

> BTW, if I remember correctly, you're pretty knowledgeable regarding
> firewire.  Do you have any comment about my STB (a DCT6200) going
> off-line?  Time Warner now has DCH3200's in my area, so I've been
> thinking of swapping it out when I add another STB.

Yeah, I know the firewire drivers and libraries reasonably well...  
Unfortunately, not well enough to know why some STB are failing to  
function. Its one of those cases where you really need the failing  
hardware in front of you, hopefully with a bus analyzer hooked up  
sniffing the traffic[1], to see exactly where things are falling down.  
And my own cable box just keeps on working without a problem.

[1] Apple's FireBug freakin' rocks for this sort of thing, though I  
really wish I could find a pcmcia/cardbus pcilynx card so I could use  
it with my powerbook...

-- 
Jarod Wilson
jarod at wilsonet.com