Hello,
On Tue, 13 Jan 2015 11:37:48 +0100, Boris Brezillon wrote:
> The DEBUG_UART_PHYS/VIRT value can only be automatically set once
> (automatically means set from the DEBUG_<SOC_ID>_<UART_ID> option), which
> prevents re-configuring the earlyprintk output for another platform once
> the DEBUG_LL output has been selected.
>
> Add a new boolean option (DEBUG_CUSTOM_UART_ADDR) enabling custom phys and
> virt address setting (configured through DEBUG_CUSTOM_UART_PHYS/VIRT
> options), and make DEBUG_UART_PHYS/VIRT hidden so that they will always be
> set from one of the 'default' definitions, and thus will be overwritten
> each time you change the DEBUG_LL output.
>
> Signed-off-by: Boris Brezillon <[email protected]>
What is the status of this patch? It would actually be quite useful
when switching back and forth between different DEBUG_LL configurations.
Russell, what do you think of the proposed approach?
Thanks,
Thomas
--
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com
On Wed, Jun 17, 2015 at 02:46:56PM +0200, Thomas Petazzoni wrote:
> What is the status of this patch? It would actually be quite useful
> when switching back and forth between different DEBUG_LL configurations.
>
> Russell, what do you think of the proposed approach?
I'd rather we killed all the defaults, rather than keep trying to fix
these subtle issues. That was always my goal of converting to this
model anyway.
However, I think it's important not to loose this information (how many
of us with lots of different platforms want to keep on digging through
PDFs to try and find this documentation?) I'd like to see a per-SoC
file somewhere in Documentation/arm/ giving these details so at least
they're relatively easy to find.
--
FTTC broadband for 0.8mile line: currently at 10.5Mbps down 400kbps up
according to speedtest.net.
Russell,
On Wed, 17 Jun 2015 20:40:58 +0100, Russell King - ARM Linux wrote:
> > Russell, what do you think of the proposed approach?
>
> I'd rather we killed all the defaults, rather than keep trying to fix
> these subtle issues. That was always my goal of converting to this
> model anyway.
>
> However, I think it's important not to loose this information (how many
> of us with lots of different platforms want to keep on digging through
> PDFs to try and find this documentation?) I'd like to see a per-SoC
> file somewhere in Documentation/arm/ giving these details so at least
> they're relatively easy to find.
I personally find having the defaults for various popular platforms
readily available in menuconfig to be a lot easier than having to dig
some documentation to find the appropriate addresses.
I.e, the current way is perfect. It's just that there is no need to
show the physical/virtual addresses if an existing well-known platform
is selected, especially as it causes weird behavior due to how kconfig
works. Which Boris patch is solving.
Best regards,
Thomas
--
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com