| Message ID | 1393264898-28939-3-git-send-email-andrew@lunn.ch (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
Subject should be "ARM: kirkwood: ..." thx, Jason. On Mon, Feb 24, 2014 at 07:01:35PM +0100, Andrew Lunn wrote: > When using platform_driver instantiation, the i2c bus was given bus > number 0. The kirkwood-t5325 audio driver has this bus number hard > coded for the address of the codec. However by default device tree i2c > busses are dynamically allocated a bus number, starting from 1. Thus > the kirkwood-t5325 cannot find its audio codec. By adding an alias in > the DT file we can control the bus number and set it to 0. The codec > can then be found. > > Signed-off-by: Andrew Lunn <andrew@lunn.ch> > --- > arch/arm/boot/dts/kirkwood.dtsi | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) > > diff --git a/arch/arm/boot/dts/kirkwood.dtsi b/arch/arm/boot/dts/kirkwood.dtsi > index 180dc69ff742..d8e34fab835b 100644 > --- a/arch/arm/boot/dts/kirkwood.dtsi > +++ b/arch/arm/boot/dts/kirkwood.dtsi > @@ -24,6 +24,7 @@ > aliases { > gpio0 = &gpio0; > gpio1 = &gpio1; > + i2c0 = &i2c0; > }; > > mbus { > @@ -111,7 +112,7 @@ > clocks = <&gate_clk 7>; > }; > > - i2c@11000 { > + i2c0: i2c@11000 { > compatible = "marvell,mv64xxx-i2c"; > reg = <0x11000 0x20>; > #address-cells = <1>; > -- > 1.8.5.3 > > > _______________________________________________ > linux-arm-kernel mailing list > linux-arm-kernel@lists.infradead.org > http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
On Mon, Feb 24, 2014 at 02:05:09PM -0500, Jason Cooper wrote:
> Subject should be "ARM: kirkwood: ..."
Humm, why?
linux/arch/arm/boot/dts$ git log --oneline armada-370.dtsi
a095b1c78a35 ARM: mvebu: sort DT nodes by address
d4fa99417a07 ARM: mvebu: link PCIe controllers to the MSI controller
238493e34d3f ARM: dts: mvebu: Update with the new compatible string for mv64xxx-
6cbe0e1feaf8 Merge tag 'dt-3.12' of git://git.infradead.org/linux-mvebu into nex
5d3b88307176 ARM: mvebu: Fix the Armada 370/XP timer compatible strings
ca60985c0042 ARM: mvebu: use correct #interrupt-cells instead of #interrupts-cel
14fd8ed0a7fd ARM: mvebu: Relocate Armada 370/XP PCIe device tree nodes
0cd3754a8317 ARM: mvebu: Add BootROM to Armada 370/XP device tree
5e12a613ce39 ARM: mvebu: Add MBus to Armada 370/XP device tree
38149887ef8b ARM: mvebu: Use the preprocessor on Armada 370/XP device tree files
9214cbe50fe7 Merge tag 'dt-3.11-5' of git://git.infradead.org/users/jcooper/linu
83735101959e ARM: mvebu: Remove device tree unused properties on A370
489e138eec96 ARM: dts: mvebu: Fix wrong the address reg value for the L2-cache n
8eed481e6fe2 arm: mvebu: fix the 'ranges' property to handle PCIe
Seems to me, now that it is part of mach-mvebu, ARM: mvebu is correct.
Andrew
Andrew, + arm-soc maintainers On Mon, Feb 24, 2014 at 08:11:08PM +0100, Andrew Lunn wrote: > On Mon, Feb 24, 2014 at 02:05:09PM -0500, Jason Cooper wrote: > > Subject should be "ARM: kirkwood: ..." > > Humm, why? > > linux/arch/arm/boot/dts$ git log --oneline armada-370.dtsi No, you're editing: arch/arm/boot/dts/kirkwood.dtsi | 3 ++- > a095b1c78a35 ARM: mvebu: sort DT nodes by address > d4fa99417a07 ARM: mvebu: link PCIe controllers to the MSI controller > 238493e34d3f ARM: dts: mvebu: Update with the new compatible string for mv64xxx- > 6cbe0e1feaf8 Merge tag 'dt-3.12' of git://git.infradead.org/linux-mvebu into nex > 5d3b88307176 ARM: mvebu: Fix the Armada 370/XP timer compatible strings > ca60985c0042 ARM: mvebu: use correct #interrupt-cells instead of #interrupts-cel > 14fd8ed0a7fd ARM: mvebu: Relocate Armada 370/XP PCIe device tree nodes > 0cd3754a8317 ARM: mvebu: Add BootROM to Armada 370/XP device tree > 5e12a613ce39 ARM: mvebu: Add MBus to Armada 370/XP device tree > 38149887ef8b ARM: mvebu: Use the preprocessor on Armada 370/XP device tree files > 9214cbe50fe7 Merge tag 'dt-3.11-5' of git://git.infradead.org/users/jcooper/linu > 83735101959e ARM: mvebu: Remove device tree unused properties on A370 > 489e138eec96 ARM: dts: mvebu: Fix wrong the address reg value for the L2-cache n > 8eed481e6fe2 arm: mvebu: fix the 'ranges' property to handle PCIe > > Seems to me, now that it is part of mach-mvebu, ARM: mvebu is correct. "ARM: kirkwood: ..." is more helpful to someone searching --oneline for changes to the kirkwood SoC. So yeah, mvebu encompasses kirkwood now with your patchset. Which makes the 'mvebu' term more vague. Before, folks could restrict commit history on mach-kirkwood/ or mach-dove/ to see changes for a particular SoC. With it all in mach-mvebu (and not in a single file), we need another way for developers to spot relevant commits. That's why I prefer the above. ex: What changed from v3.15 to v3.16 for kirkwood? $ git log --oneline v3.15..v3.16 -- arch/arm/mach-mvebu | \ grep -i 'ARM: kirkwood' This seems to be new territory for us (mvebu), perhaps one of the arm-soc maintainers has a preference? I'll add them to this reply. thx, Jason.
On Monday 24 February 2014, Jason Cooper wrote: > > ex: What changed from v3.15 to v3.16 for kirkwood? > > $ git log --oneline v3.15..v3.16 -- arch/arm/mach-mvebu | \ > grep -i 'ARM: kirkwood' > > This seems to be new territory for us (mvebu), perhaps one of the > arm-soc maintainers has a preference? I'll add them to this reply. I don't care, as long as you decide on something that works for you. While you're transitioning, it probably makes little sense to enforce either one too strictly. Arnd
On Mon, Feb 24, 2014 at 10:32:20PM +0100, Arnd Bergmann wrote: > On Monday 24 February 2014, Jason Cooper wrote: > > > > > ex: What changed from v3.15 to v3.16 for kirkwood? > > > > $ git log --oneline v3.15..v3.16 -- arch/arm/mach-mvebu | \ > > grep -i 'ARM: kirkwood' > > > > This seems to be new territory for us (mvebu), perhaps one of the > > arm-soc maintainers has a preference? I'll add them to this reply. > > I don't care, as long as you decide on something that works for you. Ok, we'll wait till we have an actual problem ;-) Seriously though, once dove and kirkwood are both migrated, we'll sit down and hash out how to handle it (if needed at all). > While you're transitioning, it probably makes little sense to > enforce either one too strictly. Agreed. Andrew, let's stick with the previous pseudo-file/path standard for now. If you're inclined, please change the dts patches to kirkwood. If you don't feel like it, I'll only twitch a little bit. :) thx, Jason.
diff --git a/arch/arm/boot/dts/kirkwood.dtsi b/arch/arm/boot/dts/kirkwood.dtsi index 180dc69ff742..d8e34fab835b 100644 --- a/arch/arm/boot/dts/kirkwood.dtsi +++ b/arch/arm/boot/dts/kirkwood.dtsi @@ -24,6 +24,7 @@ aliases { gpio0 = &gpio0; gpio1 = &gpio1; + i2c0 = &i2c0; }; mbus { @@ -111,7 +112,7 @@ clocks = <&gate_clk 7>; }; - i2c@11000 { + i2c0: i2c@11000 { compatible = "marvell,mv64xxx-i2c"; reg = <0x11000 0x20>; #address-cells = <1>;
When using platform_driver instantiation, the i2c bus was given bus number 0. The kirkwood-t5325 audio driver has this bus number hard coded for the address of the codec. However by default device tree i2c busses are dynamically allocated a bus number, starting from 1. Thus the kirkwood-t5325 cannot find its audio codec. By adding an alias in the DT file we can control the bus number and set it to 0. The codec can then be found. Signed-off-by: Andrew Lunn <andrew@lunn.ch> --- arch/arm/boot/dts/kirkwood.dtsi | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)