[meta-freescale] speed of /dev/i2c-2 in imx6q sabre-lite

[Thanassis Silis]

don;t worry about it. and thank you for all the info :)

> Date: Wed, 14 Aug 2013 09:12:52 -0700
> From: eric.nelson at boundarydevices.com
> To: djnass_18 at hotmail.com
> CC: meta-freescale at yoctoproject.org
> Subject: Re: [meta-freescale] speed of /dev/i2c-2 in imx6q sabre-lite
> 
> Hi Thanassis,
> 
> On 08/14/2013 08:40 AM, Thanassis Silis wrote:
> > Hi Eric,
> > kernel configuration is no stranger to me as I am a longtime slackware
> > user, but thank you for the info. It is always nice to know I have not
> > understood something the wrong way.
> >
> 
> Sorry if I was being pedantic. I thought that's what you were asking
> and folks come in with all kinds of background.
> 
> > What I do not know is how the IMX_I2C_BIT_RATE (which is 100kHz) is
> > related to the imxi2c_platform_data structure.
> > The file drivers/i2c/busses/i2c_imx.c doesn't include
> > arch/arm/mach-mx6/board_nitrogen6x.c (or _sabrelite.c for that matter)
> > or vice-versa. So is it correct to say that i2c-imx.c sets the bitrate
> > earlier in the boot process and then the platform specific code,
> > overwrites/alters that value to what is set in board_nitrogen6x.c ?
> 
> Nope. The driver's "probe" routine gets handed the imxi2c_platform_data
> structure from the device registration procedure in
> board-mx6_nitrogen.c.
> 
> IOW, there's a run-time path, not a compile-time path. Otherwise, every
> driver would need to know about every board that used it.
> 
> If you look closely, all of the uses of IMX_I2C_BIT_RATE are conditional
> on either missing "pdata" (platform data) or a zero value for bitrate.
> 
> Regards,
> 
> 
> Eric
 		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.yoctoproject.org/pipermail/meta-freescale/attachments/20130814/c1a866fa/attachment.html>