> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]] On Behalf
> Of Subrahmanya, Chaithrika
> Sent: Monday, January 12, 2009 5:14 PM
> To: Kevin Hilman; [email protected]
> Subject: RE: davinci git updated to v2.6.28
> 
> > -----Original Message-----
> > From: davinci-linux-open-source-
> > [email protected] [mailto:davinci-linux-
> > [email protected]] On Behalf
> Of
> > Kevin Hilman
> > Sent: Saturday, January 10, 2009 1:28 PM
> > To: [email protected]
> > Subject: Re: davinci git updated to v2.6.28
> >
> > Kevin Hilman <[email protected]> writes:
> >
> > > DaVinci git has been sync'd with the upstream (omap) tree at
> > > v2.6.28.  Boot tested a single kernel (davinci_all_defconfig) on
> > > dm6446, dm355 and dm6467.
> >
> > FYI...
> >
> > I boot tested on dm6467, but I just now noticed that when using
> > CONFIG_DEBUG_LL on dm6467 I'm back to getting garbage out the serial
> > port.  dm355 and dm6446 work fine with CONFIG_DEBUG_LL.
> >
> > I haven't yet debugged this further.
> 
> This is due to the byte access of UART through the macros in
> arch/arm/include/asm/hardware/debug-8250.S file. On changing the
> instructions to
> 'ldr' and 'str' from 'ldrb' and 'strb' respectively, the debug messages
> are output.
> 

'DM646x requires 32-bit access to UART unlike other DaVinci SoCs'.
The changes made to arch/arm/include/asm/hardware/debug-8250.S file
was just to test and a better way to override this behavior is still being 
looked into.

-Chaithrika_______________________________________________
Davinci-linux-open-source mailing list
[email protected]
http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source

Reply via email to