Hello, I have a question. I see in the sources there are drivers for i2c devices. What I am thinking is to write a driver to program and initialize the tv-out chip through the i2c (smbus). The tv-out chip is embedded in the board and programable through i2c. If this is possible I would need to know at what point the video bios gets initialized. And I would want to load this driver right after video bios gets initialized correct? Thoughts?
Thanks - Joe
hey Joe,
I'd guess, during testing, the sequence would make not much difference. Once the code is working, the earlier it gets called, the more you could output onto TV which would be desirable. maybe use it as debug output device. greetinx: --Q
joe@smittys.pointclark.net schrieb:
Hello, I have a question. I see in the sources there are drivers for i2c devices. What I am thinking is to write a driver to program and initialize the tv-out chip through the i2c (smbus). The tv-out chip is embedded in the board and programable through i2c. If this is possible I would need to know at what point the video bios gets initialized. And I would want to load this driver right after video bios gets initialized correct? Thoughts?
On Tue, Mar 13, 2007 at 04:50:47AM +0100, Quux wrote:
hey Joe,
I'd guess, during testing, the sequence would make not much difference. Once the code is working, the earlier it gets called, the more you could output onto TV which would be desirable. maybe use it as debug output device. greetinx: --Q
joe@smittys.pointclark.net schrieb:
Hello, I have a question. I see in the sources there are drivers for i2c devices. What I am thinking is to write a driver to program and initialize the tv-out chip through the i2c (smbus). The tv-out chip is embedded in the board and programable through i2c. If this is possible I would need to know at what point the video bios gets initialized. And I would want to load this driver right after video bios gets initialized correct? Thoughts?
TV encoders are not standalone devices. They get pixel data from your graphics device, and they either provide or receive sync signals, blanking and clocking either to or from the graphics device. It doesn't matter if these devices are programmed over smbus or over graphics device local I2C/GPIO lines.
So this is an integral part part of graphics device initialisation. As such, it is all up to the VGA BIOS, or, if available, up to the direct support of certain graphics devices.
Luc Verhaegen. http://unichrome.sf.net/
Quoting Luc Verhaegen libv@skynet.be:
On Tue, Mar 13, 2007 at 04:50:47AM +0100, Quux wrote:
hey Joe,
I'd guess, during testing, the sequence would make not much difference. Once the code is working, the earlier it gets called, the more you could output onto TV which would be desirable. maybe use it as debug output device. greetinx: --Q
joe@smittys.pointclark.net schrieb:
Hello, I have a question. I see in the sources there are drivers for i2c devices. What I am thinking is to write a driver to program and initialize the tv-out chip through the i2c (smbus). The tv-out chip is embedded in the board and programable through i2c. If this is possible I would need to know at what point the video bios gets initialized. And I would want to load this driver right after video bios gets initialized correct? Thoughts?
TV encoders are not standalone devices. They get pixel data from your graphics device, and they either provide or receive sync signals, blanking and clocking either to or from the graphics device. It doesn't matter if these devices are programmed over smbus or over graphics device local I2C/GPIO lines.
So this is an integral part part of graphics device initialisation. As such, it is all up to the VGA BIOS, or, if available, up to the direct support of certain graphics devices.
Luc Verhaegen. http://unichrome.sf.net/
I understand video encoders are not standalone. They receive input from your GPU, in my case it is embedded so the northbridge, encode it, and send it out the output. I have a vga bios that outputs to DVOA. This is the the port that is connected to the input of the video encoder. Unfortunatly (intel) never released a module for this video encoder. So, this is what I want to do. Write a driver to initialize the video encoder and then when the vga bios starts outputting to DVOA we should have signal correct?
Thanks - Joe
On Tue, Mar 13, 2007 at 12:57:27AM -0400, joe@smittys.pointclark.net wrote:
Quoting Luc Verhaegen libv@skynet.be:
On Tue, Mar 13, 2007 at 04:50:47AM +0100, Quux wrote:
I understand video encoders are not standalone. They receive input from your GPU, in my case it is embedded so the northbridge, encode it, and send it out the output. I have a vga bios that outputs to DVOA. This is the the port that is connected to the input of the video encoder. Unfortunatly (intel) never released a module for this video encoder. So, this is what I want to do. Write a driver to initialize the video encoder and then when the vga bios starts outputting to DVOA we should have signal correct?
Thanks - Joe
Your only real chance is doing this shortly _after_ the VGA BIOS has run. That way you can adjust to whatever tricks the VGA BIOS pulls.
Luc Verhaegen.
Yeh, As soon as I get my setup working, I think I am going to call the tv-out driver just after the superIO initializes this way when ever the vga bios starts it's output to DVOA the tv-out chip will be ready to encode.
Thanks - Joe
Quoting Quux pawn2be.wild@yahoo.de:
hey Joe,
I'd guess, during testing, the sequence would make not much difference. Once the code is working, the earlier it gets called, the more you could output onto TV which would be desirable. maybe use it as debug output device. greetinx: --Q
joe@smittys.pointclark.net schrieb:
Hello, I have a question. I see in the sources there are drivers for i2c devices. What I am thinking is to write a driver to program and initialize the tv-out chip through the i2c (smbus). The tv-out chip is embedded in the board and programable through i2c. If this is possible I would need to know at what point the video bios gets initialized. And I would want to load this driver right after video bios gets initialized correct? Thoughts?
-- linuxbios mailing list linuxbios@linuxbios.org http://www.openbios.org/mailman/listinfo/linuxbios
-
joe＠smittys.pointclark.net -
Luc Verhaegen -
Quux