Re: [PATCH 0/6] ARM: at91: irqdomain and device tree for AIC and GPIO
From: Nicolas Ferre <hidden>
Date: 2011-12-16 16:29:29
Also in:
linux-arm-kernel
On 12/16/2011 02:53 AM, Rob Herring :
Nicolas, On 12/15/2011 01:16 PM, Nicolas Ferre wrote:quoted
Hi, This series adds irqdomain and device tree support for both the interrupt and GPIO controllers of AT91 SoC. The AIC part has already been sent some time ago but has been reworked to address Rob's comments. This reworked patch is marked with a "v4" tag. The series can go on top of Jamie's patch "irqdomain: export irq_domain_simple_ops for !CONFIG_OF" and has been tested on top of 'v3.2-rc5' + 'arm-soc/at91/ioremap' and 'arm-soc/at91/gpio' Jean-Christophe PLAGNIOL-VILLARD (1): ARM: at91/gpio: add DT support Nicolas Ferre (5): ARM: at91/aic: add irq domain and device tree support ARM: at91/gpio: add irqdomain to gpio interrupts ARM: at91/gpio: add .to_irq gpio_chip handler and rework irq_to_gpio ARM: at91/gpio: remove the static specification of gpio_chip.base ARM: at91/board-dt: remove AIC irq domain from board file .../devicetree/bindings/arm/atmel-aic.txt | 38 ++++ .../devicetree/bindings/gpio/gpio_at91.txt | 20 ++ arch/arm/Kconfig | 1 + arch/arm/boot/dts/at91sam9g20.dtsi | 44 ++++- arch/arm/boot/dts/at91sam9g45.dtsi | 59 +++++- arch/arm/mach-at91/board-dt.c | 15 +-- arch/arm/mach-at91/gpio.c | 197 ++++++++++++++++---- arch/arm/mach-at91/include/mach/gpio.h | 12 -- arch/arm/mach-at91/include/mach/irqs.h | 3 +- arch/arm/mach-at91/irq.c | 91 +++++++-- 10 files changed, 383 insertions(+), 97 deletions(-) create mode 100644 Documentation/devicetree/bindings/arm/atmel-aic.txt create mode 100644 Documentation/devicetree/bindings/gpio/gpio_at91.txtWhat about using generic irq chip as I previously mentioned? It looks to me like both the AIC and gpio controller could use it. I've added irq domain support to generic irq chip, so it should simplify your changes. The latest branch is here: git://sources.calxeda.com/kernel/linux.git pl061-domain-v2
Rob, Yes, that definitively makes sense but I fear that this rework will delay again the introduction of a couple of products and the work we have to do to enable DT for other drivers. The AIC and GPIO will definitively need a rework but I cannot start this now and as I have a nearly working solution, I would prefer to start this task later... Best regards, -- Nicolas Ferre