lkml.org 
[lkml]   [2014]   [Jan]   [2]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
SubjectRe: [PATCH 6/6] at91: dt: sam9261: Added DM9000 in the device tree
From
2013/12/31 Arnd Bergmann <arnd@arndb.de>:
> On Tuesday 31 December 2013 17:32:14 jjhiblot@traphandler.com wrote:
>>
>> + ethernet@30000000 {
>> + compatible = "davicom,dm9000";
>> + reg = <0x30000000 0x2 0x30000004 0x2>;
>> + interrupt-parent = <&pioC>;
>> + interrupts = <11 IRQ_TYPE_EDGE_BOTH>;
>> + local-mac-address = [00 00 de ad be ef];
>> + davicom,no-eeprom;
>> +
>> + smc,cs = <2>;
>> + smc,ncs_read_setup = <0>;
>> + smc,nrd_setup = <2>;
>> + smc,ncs_write_setup = <0>;
>> + smc,nwe_setup = <2>;
>> + smc,ncs_read_pulse = <8>;
>> + smc,nrd_pulse = <4>;
>> + smc,ncs_write_pulse = <8>;
>> + smc,nwe_pulse = <4>;
>> + smc,read_cycle = <16>;
>> + smc,write_cycle = <16>;
>> + smc,tdf_cycles = <1>;
>> + smc,tdf_optimized = <0>;
>> + smc,page_size = <0>;
>> + smc,byte_access_type = <1>;
>> + smc,bus_width = <1>;
>> + smc,nwait_mode = <0>;
>> + smc,read_mode = <1>;
>> + smc,write_mode = <1>;
>> + };
>
> I don't like how this forces you to mix the properties
> of the ethernet device with the properties of the smc
> port. I would also like to see the address translation
> reflected in the DT nodes. This should probably just work
> out if you restructure the DT representation, and won't need
> changes in the driver.
>
> How about this:
>
> smc: smc@ffffec00 {
> #address-cells = <2>; /* cs, address */
> #size-cells = <1>;
> ranges = <2 0 0x30000000 0x10000000>
> <3 0 0x40000000 0x10000000>;
>
> extbus@2.0 {
> compatible = "simple-bus";
> #address-cells = <1>;
> #size-cells = <1>;
> ranges = <0 2 0 0x10000000>;
> smc,ncs_read_setup = <0>;
> smc,nrd_setup = <2>;
> smc,...
>
> ethernet@0 {
> compatible = "davicom,dm9000";
> reg = <0x0 0x2> <0x4 0x2>;
> };
> };
> };
I'm okay with this. I'll do something along those lines in the next
round of the patch set.
>
> If each bus interface has a fixed range of registers, you can actually
> collapse the upper two nodes again and just have one device per
> chipselect.
This is indeed the case, but I'd rather keep the top node to highlight
the fact that all the devices are physically tied to the same bus.

Jean-Jacques

>
> Arnd
>
> _______________________________________________
> linux-arm-kernel mailing list
> linux-arm-kernel@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel


\
 
 \ /
  Last update: 2014-01-03 00:21    [W:0.133 / U:0.816 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site