ivtv-devel

Steve,

On Sun, 2009-06-21 at 11:02 +0100, steve wrote:
> Andy,
>       I tried the patches for the ACER 500 that you posted. 
> 
> Purely FYI There was just one small problem with the compile in that my
> version of cx18-drivers.h was missing the LEADTEK_DVR1300H entry
> which I added as below:
> 
> 
> #define CX18_CARD_LEADTEK_PVR2100     6 /* Leadtek WinFast
> PVR2100/DVR3100 H */
> #define CX18_CARD_LEADTEK_DVR3100H    7 /* Leadtek WinFast
> PVR2100/DVR3100 H */
> #define CX18_CARD_LAST                7
> 
> 
>  I probably have an old version. 

You may eventually want a clone or tarball of my working repository for
the MPC-718.

For a tarball:

http://linuxtv.org/hg/~awalls/cx18-mpc718/archive/tip.tar.bz2

For a local working repository copy:

hg clone http://linuxtv.org/hg/~awalls/cx18-mpc718

Or to stick with what you have and pick up new patches individually:

http://linuxtv.org/hg/~awalls/cx18-mpc718/



There were some changes made to the cx18 driver and xc2028 tuner driver
to get the DVR3100H working with DVB-T (those changes are in my
repository and also in the latest v4l-dvb).  I think for 8 MHz UHF
DVB-T, you'll be OK even if you stick with just applying the specific
patches I made for the Yuan MPC-718.


> Thereafter it compiled and installed,
> but I no longer get any /dev/video* (which I did get previously)
> nor /dev/dvb/* (which I did NOT get previously). I've posted
> the relevant section from dmesg below. This was following 

Looking below, it appears that I guessed the wrong I2C address for the
XC3028 tuner chip.  This caused DVB registration to fail and hence
driver init to fail.  I have added a patch to use the correct I2C
address.



>  adding patch .c files to directory
>  sudo make 
>  sudo make install
>  "reboot"
> 
>  I've not seen the "bad EEPROM" message before.

That's because I added support for reading the EEPROM that I saw on the
MPC-718 (it's on the back).  Either I goofed something up, or you didn't
get the patch applied properly to cx18-driver.c:cx18_read_eeprom() and
cx18-driver.c:cx18_process_eeprom.

You should have seen an eeprom dump and then the message:

        "Unknown EEPROM encoding"

Either way, things should have proceeded anyway.

I have added a patch to disable trying to obtain an EEPROM dump from the
MPC718 (for now).


> Please also find below lspci output, in case it helps.
> 
> Please note that I currently only have one card installed, the second
> one is still on the bench in case you need any more hardware data, I
> don't want to risk damaging it with too much unnecessary soldering
> and removals.

That's fine.  The only trip-up would be a ZL10353 chip being installed
on the new card instead of a MT352 chip.

I've added a patch to hopefully handle an MPC-718 with a ZL10353 as
well.


> dmesg
> 
> [   11.942443] cx18:  Start initialization, version 1.2.0
> [   11.942502] cx18-0: Initializing card 0
> [   11.942505] cx18-0: Autodetected Yuan MPC718 MiniPCI DVB-T/Analog
> card

Good.

> [   11.943341] cx18 0000:02:00.0: PCI INT A -> GSI 20 (level, low) ->
> IRQ 20
> [   11.943352] cx18-0: Unreasonably low latency timer, setting to 64
> (was 2)
> [   11.946434] cx18-0: cx23418 revision 01010000 (B)
> [   11.949211] phy0: Selected rate control algorithm 'pid'
> [   12.078412] Broadcom 43xx driver loaded [ Features: PLR, Firmware-ID:
> FW13 ]
> [   12.102583] HDA Intel 0000:00:1b.0: PCI INT A -> GSI 16 (level, low)
> -> IRQ 16
> [   12.102639] HDA Intel 0000:00:1b.0: setting latency timer to 64
> [   12.125858] tveeprom 0-0050: Encountered bad packet header [00].
> Corrupt or not a Hauppauge eeprom.
> [   12.125863] cx18-0: Invalid EEPROM

Hmmm.  This should have been an eeprom dump with an "Unknown EEPROM
encoding" message.  But the driver will ignore this anyway.

> [   12.125867] cx18-0: Experimenters needed for device to work well.
> [   12.125869]        To help, mail the ivtv-devel list (www.ivtvdriver.org).
> [   12.258180] tuner 1-0061: chip found @ 0xc2 (cx18 i2c driver #0-1)
> [   12.306617] xc2028 1-0061: creating new instance
> [   12.306620] xc2028 1-0061: type set to XCeive xc2028/xc3028 tuner
                          ^^^^
Ooops, I guessed address 0x60 not 0x61.  This is likely why DVB
registration failed below.

> [   12.306986] cx18-0: Registered device video0 for encoder MPEG (64 x
> 32 kB)
> [   12.306990] DVB: registering new adapter (cx18)
> [   12.307305] cx18-0: frontend initialization failed
> [   12.307508] cx18-0: DVB failed to register

Likely due to using the wrong I2C address for the XC3028 tuner.


> [   12.307570] cx18-0: Registered device video32 for encoder YUV (16 x
> 128 kB)
> [   12.307628] cx18-0: Registered device vbi0 for encoder VBI (20 x
> 51984 bytes)
> [   12.307688] cx18-0: Registered device video24 for encoder PCM audio
> (256 x 4 kB)
> [   12.307763] cx18-0: Registered device radio0 for encoder radio
> [   12.308241] cx18-0: Error -1 registering devices
> [   12.308320] xc2028 1-0061: destroying instance
> [   12.308892] cx18-0: Error -1 on initialization
> [   12.308902] cx18: probe of 0000:02:00.0 failed with error -1
> [   12.308952] cx18:  End initialization

If DVB fails, everything fails. :(

I'll have to change that someday.


> [EMAIL PROTECTED]:~$ ls /dev/dvb*
> ls: cannot access /dev/dvb*: No such file or directory
> [EMAIL PROTECTED]:~$ ls /dev/vid*
> ls: cannot access /dev/vid*: No such file or directory


Thanks for the testing and feedback.  PLease continue testing.  BTW you
may want to add a

 options cx18 debug=255
 options tuner-xc2028 debug=3
 options mt352 debug=3
 options zl10353 debug=3
 options tuner debug=3

to your /etc/modprobe.conf  (or equivalent file) to have the drivers log
somewhat verbosely while figuring things out.  (I picked debug level '3'
out of the air for the non-cx18 modules.  That should give enough debug
level for now.)

Regards,
Andy


_______________________________________________
ivtv-devel mailing list
[EMAIL PROTECTED]
http://ivtvdriver.org/mailman/listinfo/ivtv-devel