On Wednesday 24 January 2007 19:35, Roger James wrote:
> > -----Original Message-----
> > From: [EMAIL PROTECTED] [mailto:ivtv-users-
> > [EMAIL PROTECTED] On Behalf Of Roger James
> > Sent: 24 January 2007 00:24
> > To: 'User discussion about IVTV'
> > Subject: Re: [ivtv-users] saa7115 not autoloading on 2.6.19.2
> > kernel
> >
> > > On Tuesday 23 January 2007 23:05, Roger James wrote:
> > > > Comments inline
> > > >
> > > > > On Tuesday 23 January 2007 19:46, Roger James wrote:
> > > > > > I know this has been mentioned before, but I cannot get it
> > > > > > to work. I have just moved to kernel version 2.6.19.2
> > > > > > ivtv-0.9.1 and the latest v4l-dvb (hg). I followed the
> > > > > > instructions in the readme and ensured that I updated the
> > > > > > kernel config to build the various support drivers
> > > > > > including the saa7115. But I still get the "i2c_adapter
> > > > > > i2c-3: Client creation failed at 0x44 (-1)" message in the
> > > > > > ivtv init.
> > > > >
> > > > > This message is not from the saa7115 module, it's some other
> > > > > module that's giving you this message. I get the same, but it
> > > > > is harmless.
> > > >
> > > > The message comes from the i2c_core module and appears to be
> > > > that module reporting that the ivtv modules request to attach
> > > > the saa7115 driver (0x44) has failed. In my case this is not
> > > > benign as an lsmod afterwards shows that the saa7115 driver
> > > > module has not been loaded.
> > >
> > > Hmm. Can you post the INIT IVTV messages in the kernel log?
> > >
> > > Hans
> >
<snip>
>
> Some more info on this. I now suspect that 0x44 is the I2C address of
> saa7127 driver (see below) which is also not being loaded.
Yes, I can confirm that on my system as well. But I only see that on
non-PVR cards where the saa7127 driver (correctly) detects that it
isn't a saa7127 at that address after all.
> I also get the following message further down the log, which I think
> is telling me that the saa7115 (address 0x21 is driver is not being
> found.
>
> Jan 24 17:37:14 myth kernel: ivtv0: i2c addr 0x21 not found for
> command 0x4008646f!
Weird.
> [EMAIL PROTECTED]:~/ivtv-0.9.1/driver$ vi ivtv-i2c.c
> #endif /* I2C_ADAP_CLASS_TV_ANALOG */
>
> #define IVTV_CS53L32A_I2C_ADDR 0x11
> #define IVTV_CX25840_I2C_ADDR 0x44
> #define IVTV_SAA7115_I2C_ADDR 0x21
> #define IVTV_SAA7127_I2C_ADDR 0x44
> #define IVTV_SAA717x_I2C_ADDR 0x21
> #define IVTV_MSP3400_I2C_ADDR 0x40
> #define IVTV_HAUPPAUGE_I2C_ADDR 0x50
> #define IVTV_WM8739_I2C_ADDR 0x1a
> #define IVTV_WM8775_I2C_ADDR 0x1b
> #define IVTV_TEA5767_I2C_ADDR 0x60
> #define IVTV_UPD64031A_I2C_ADDR 0x12
> #define IVTV_UPD64083_I2C_ADDR 0x5c
> #define IVTV_TDA985X_I2C_ADDR 0x5b
>
> So it seems to come back to my original question. Who is responsible
> for loading these modules and how should it be made to happen.
ivtv loads these modules. The ivtv_request_module() function in
ivtv-driver.c loads each module.
There are two possible reasons why this might fail:
1) you may need to run 'depmod' to update depencies
2) when ivtv is built it checks kernel config defines to see whether a
driver is compiled as a module or in-kernel. It won't load anything if
it is in-kernel. So if there is a mismatch between your kernel and the
config header then that too might be a cause for failure.
I'm convinced it is something in your environment, since it is the first
time anyone has ever complained about this.
>
> I seem to remember in the dim and distant past you had to list them
> as dependencies in the modutils config.
That was needed for very old drivers, but not since a long time.
Regards,
Hans
>
> Roger
>
>
>
> _______________________________________________
> ivtv-users mailing list
> [email protected]
> http://ivtvdriver.org/mailman/listinfo/ivtv-users
_______________________________________________
ivtv-users mailing list
[email protected]
http://ivtvdriver.org/mailman/listinfo/ivtv-users
