> Date: Sat, 24 May 2008 09:07:15 -0400 > From: Andy Walls > Subject: Re: [ivtv-users] CX18: What Next? > To: User discussion about IVTV > Message-ID: <[EMAIL PROTECTED]> > Content-Type: text/plain; charset="us-ascii" > > On Thu, 2008-05-22 at 17:25 -0700, Michael wrote: > > > On Wed, 2008-05-21 at 17:09 -0700, Michael wrote: > > > > > > > > > > > > Could you post the output from /var/log/messages instead of from dmesg? > > > dmesg has a finite amount of space, and the beginning of the > > > initialization was lost. The beginning's the part I really need to see, > > > where the i2c bus is being initialized and reset and the EEPROM is being > > > addressed and read. > > > > > Here you go. > > > > http://h1.ripway.com/hendrick/log.txt > > > > EEPROM is still there. > > >From the I2C debug in your latest log, it's highly likely there is a > device other than the EEPROM that is stuck holding the SDA line low. > > > Also, I see in your latest log: > > May 22 20:16:46 localhost kernel: cx18-0 info: GPIO initial dir: > 0000ffff/0000ffff out: 00000000/00000000 > > > Which is a dump of the GPIO line state before the cx18 driver messes > with them. This matches my card at boot up (with debug=3 in > modprobe.conf): > > May 24 07:03:01 palomino kernel: cx18-0 info: GPIO initial dir: > 0000ffff/0000ffff out: 00000000/00000000 > > > For me, after one modprobe -r cx18; modprobe cx18 debug=3 sequence, I have: > > May 24 07:05:35 palomino kernel: cx18-0 info: GPIO initial dir: > 0000cffe/0000ffff out: 00003001/00000000 > > reflecting that the initial cx18 driver load raised some GPIO lines and > left them raised. > > Comments, in cx18-gpio.c indicate that these are the lines that reset > the Conexant CX24227, Zilog Z8F0811, and Cirrus Logic CS5345 - all > devices that could be stuck holding the SDA line of the I2C bus low. > > > > Attached is a patch to actively reset these I2C slave chips via GPIO > upon modprobe, before the EEPROM is interrogated. The patch is not > perfect, but it should be OK for systems with only one HVR-1600 card > installed. It does no harm to my setup. The patch can be applied with > or without the previous two patches I provided to you. > > Let me know if this enables the EEPROM to be detected reliably. >
I would like to apologize first off for the slow response. I somehow missed this message in my inbox, and when I went back to check today, I found it. When I run make, with this patch applied, I get the following error: CC [M] /home/michael/Programs/TVTuner/v4l-dvb-bef996fa0ded/v4l/cx18-gpio.o /home/michael/Programs/TVTuner/v4l-dvb-bef996fa0ded/v4l/cx18-gpio.c: In function 'cx18_gpio_i2c_slave_reset': /home/michael/Programs/TVTuner/v4l-dvb-bef996fa0ded/v4l/cx18-gpio.c:73: error: 'gpio_val' undeclared (first use in this function) /home/michael/Programs/TVTuner/v4l-dvb-bef996fa0ded/v4l/cx18-gpio.c:73: error: (Each undeclared identifier is reported only once /home/michael/Programs/TVTuner/v4l-dvb-bef996fa0ded/v4l/cx18-gpio.c:73: error: for each function it appears in.) make[3]: *** [/home/michael/Programs/TVTuner/v4l-dvb-bef996fa0ded/v4l/cx18-gpio.o] Error 1 make[2]: *** [_module_/home/michael/Programs/TVTuner/v4l-dvb-bef996fa0ded/v4l] Error 2 make[2]: Leaving directory `/usr/src/linux-2.6.22.17.tex2' make[1]: *** [default] Error 2 make[1]: Leaving directory `/home/michael/Programs/TVTuner/v4l-dvb-bef996fa0ded/v4l' make: *** [all] Error 2 It seems to compile fine without the patch. I tried compiling this without the other two patches, if you want, I'll find those patches, and try again. Thanks again for the help, and I apologize once again for the slow response. > > Regards, > Andy > > -------------- next part -------------- > A non-text attachment was scrubbed... > Name: cx18-i2c-gpio-reset.patch > Type: text/x-patch > Size: 3397 bytes > Desc: not available > Url : > http://ivtvdriver.org/pipermail/ivtv-users/attachments/20080524/a6742c48/attachment-0001.bin > _______________________________________________ ivtv-users mailing list [email protected] http://ivtvdriver.org/mailman/listinfo/ivtv-users
