[PATCH v2 0/8] Second patchset for LPC32xx device tree conversion

[PATCH v2 0/8] Second patchset for LPC32xx device tree conversion

[Roland Stigge]

This is the second patch set for device tree conversion of mach-lpc32xx. It
builds upon and requires the previously posted first set of 8 plus the GPIO
patch.

It touches different subsystems (USB/OHCI, input/touchscreen, iio/adc and
mach-lpc32) to be picked by the respective subsystem maintainer and probably
applied separately. (?)

Signed-off-by: Roland Stigge <stigge@antcom.de>

--

Applies to v3.4-rc3

Changes since v1:
* Separated out DTS files
* Separated out clock registration update
* Separated out removal of obsolete Kconfig
ohci-nxp.c:
* Separated out cleanup patch from device tree conversion
* Cleanup of resource handling (mem and irq), use devm_request_and_ioremap()
* Fixed controller initialization (bits)
* Use DMA_BIT_MASK()
* Don't remove __devinit
* Removed obsolete gpio.h include
* More consistent bit access symbols
* Removed unused extern declaration ocpi_enable()
* Added error handling on i2c_get_adapter()
* Better USB host and device differentiation, since on LPC32xx, the same
  controller can be used as USB device controller
* Whitespace cleanup

Thanks to Thierry Reding for reviewing!

You can also pull from:

	git://git.antcom.de/linux-2.6 lpc32xx/dt2

Roland Stigge (8):
ohci-nxp: Driver cleanup
ohci-nxp: Device tree support
ARM: LPC32xx: clock.c update
ARM: LPC32xx: Remove obsolete platform Kconfig
input: Device tree support for LPC32xx touchscreen
iio: lpc32xx-adc: Remove driver conflict due to device tree
ARM: LPC32xx: DTS files for device tree conversion
ARM: LPC32xx: Device tree support

 Documentation/devicetree/bindings/arm/lpc32xx-mic.txt               |   38 +
 Documentation/devicetree/bindings/arm/lpc32xx.txt                   |    8 
 Documentation/devicetree/bindings/input/touchscreen/lpc32xx-tsc.txt |   16 
 Documentation/devicetree/bindings/usb/ohci-nxp.txt                  |   16 
 arch/arm/Kconfig                                                    |    1 
 arch/arm/boot/dts/lpc32xx.dtsi                                      |  303 ++++++++++
 arch/arm/boot/dts/phy3250.dts                                       |  115 +++
 arch/arm/mach-lpc32xx/Kconfig                                       |   26 
 arch/arm/mach-lpc32xx/clock.c                                       |   77 +-
 arch/arm/mach-lpc32xx/common.c                                      |  192 ------
 arch/arm/mach-lpc32xx/common.h                                      |   14 
 arch/arm/mach-lpc32xx/irq.c                                         |   78 +-
 arch/arm/mach-lpc32xx/phy3250.c                                     |  146 ++--
 drivers/input/touchscreen/lpc32xx_ts.c                              |   10 
 drivers/staging/iio/adc/Kconfig                                     |    8 
 drivers/usb/host/ohci-nxp.c                                         |   55 +
 16 files changed, 704 insertions(+), 399 deletions(-)

[PATCH v2 1/8] ohci-nxp: Driver cleanup

[Roland Stigge]

Cleanup for ohci-nxp.c:

* Cleanup of resource handling (mem and irq), use devm_request_and_ioremap()
* Fixed controller initialization (bits)
* Use DMA_BIT_MASK()
* Don't remove __devinit
* Removed obsolete gpio.h include
* More consistent bit access symbols
* Removed unused extern declaration ocpi_enable()
* Added error handling on i2c_get_adapter()
* Better USB host and device differentiation, since on LPC32xx, the same
  controller can be used as USB device controller
* Whitespace cleanup

Signed-off-by: Roland Stigge <stigge@antcom.de>

---

Applies to v3.4-rc3

Thanks to Thierry Reding for reviewing!

 drivers/usb/host/ohci-nxp.c |   45 ++++++++++++++++++++++++++------------------
 1 file changed, 27 insertions(+), 18 deletions(-)

--- linux-2.6.orig/drivers/usb/host/ohci-nxp.c
+++ linux-2.6/drivers/usb/host/ohci-nxp.c
@@ -29,7 +29,6 @@
 
 #include <mach/platform.h>
 #include <mach/irqs.h>
-#include <asm/gpio.h>
 
 #define USB_CONFIG_BASE		0x31020000
 #define PWRMAN_BASE		0x40004000
@@ -38,7 +37,9 @@
 
 /* USB_CTRL bit defines */
 #define USB_SLAVE_HCLK_EN	(1 << 24)
+#define USB_DEV_NEED_CLK_EN	(1 << 22)
 #define USB_HOST_NEED_CLK_EN	(1 << 21)
+#define PAD_CONTROL_LAST_DRIVEN	(1 << 19)
 
 #define USB_OTG_CLK_CTRL	IO_ADDRESS(USB_CONFIG_BASE + 0xFF4)
 #define USB_OTG_CLK_STAT	IO_ADDRESS(USB_CONFIG_BASE + 0xFF8)
@@ -117,7 +118,6 @@ static struct i2c_driver isp1301_driver;
 static struct i2c_client *isp1301_i2c_client;
 
 extern int usb_disabled(void);
-extern int ocpi_enable(void);
 
 static struct clk *usb_clk;
 
@@ -220,7 +220,7 @@ static void isp1301_configure_lpc32xx(vo
 		ISP1301_I2C_INTERRUPT_RISING | ISP1301_I2C_REG_CLEAR_ADDR, ~0);
 
 	/* Enable usb_need_clk clock after transceiver is initialized */
-	__raw_writel((__raw_readl(USB_CTRL) | (1 << 22)), USB_CTRL);
+	__raw_writel((__raw_readl(USB_CTRL) | USB_HOST_NEED_CLK_EN), USB_CTRL);
 
 	printk(KERN_INFO "ISP1301 Vendor ID  : 0x%04x\n",
 	      i2c_smbus_read_word_data(isp1301_i2c_client, 0x00));
@@ -374,9 +374,13 @@ static int __devinit usb_hcd_nxp_probe(s
 	const struct hc_driver *driver = &ohci_nxp_hc_driver;
 	struct i2c_adapter *i2c_adap;
 	struct i2c_board_info i2c_info;
+	struct resource *res;
 
 	int ret = 0, irq;
 
+	pdev->dev.coherent_dma_mask = DMA_BIT_MASK(32);
+	pdev->dev.dma_mask = &pdev->dev.coherent_dma_mask;
+
 	dev_dbg(&pdev->dev, "%s: " DRIVER_DESC " (nxp)\n", hcd_name);
 	if (usb_disabled()) {
 		err("USB is disabled");
@@ -384,16 +388,8 @@ static int __devinit usb_hcd_nxp_probe(s
 		goto out;
 	}
 
-	if (pdev->num_resources != 2
-	    || pdev->resource[0].flags != IORESOURCE_MEM
-	    || pdev->resource[1].flags != IORESOURCE_IRQ) {
-		err("Invalid resource configuration");
-		ret = -ENODEV;
-		goto out;
-	}
-
 	/* Enable AHB slave USB clock, needed for further USB clock control */
-	__raw_writel(USB_SLAVE_HCLK_EN | (1 << 19), USB_CTRL);
+	__raw_writel(USB_SLAVE_HCLK_EN | PAD_CONTROL_LAST_DRIVEN, USB_CTRL);
 
 	ret = i2c_add_driver(&isp1301_driver);
 	if (ret < 0) {
@@ -401,6 +397,11 @@ static int __devinit usb_hcd_nxp_probe(s
 		goto out;
 	}
 	i2c_adap = i2c_get_adapter(2);
+	if (!i2c_adap) {
+		err("failed on i2c_get_adapter");
+		goto out_i2c_driver;
+	}
+
 	memset(&i2c_info, 0, sizeof(struct i2c_board_info));
 	strlcpy(i2c_info.type, "isp1301_nxp", I2C_NAME_SIZE);
 	isp1301_i2c_client = i2c_new_probed_device(i2c_adap, &i2c_info,
@@ -435,6 +436,7 @@ static int __devinit usb_hcd_nxp_probe(s
 	}
 
 	__raw_writel(__raw_readl(USB_CTRL) | USB_HOST_NEED_CLK_EN, USB_CTRL);
+	__raw_writel(__raw_readl(USB_CTRL) & ~USB_DEV_NEED_CLK_EN, USB_CTRL);
 
 	/* Set to enable all needed USB clocks */
 	__raw_writel(USB_CLOCK_MASK, USB_OTG_CLK_CTRL);
@@ -442,7 +444,7 @@ static int __devinit usb_hcd_nxp_probe(s
 	while ((__raw_readl(USB_OTG_CLK_STAT) & USB_CLOCK_MASK) !=
 	       USB_CLOCK_MASK) ;
 
-	hcd = usb_create_hcd (driver, &pdev->dev, dev_name(&pdev->dev));
+	hcd = usb_create_hcd(driver, &pdev->dev, dev_name(&pdev->dev));
 	if (!hcd) {
 		err("Failed to allocate HC buffer");
 		ret = -ENOMEM;
@@ -452,14 +454,21 @@ static int __devinit usb_hcd_nxp_probe(s
 	/* Set all USB bits in the Start Enable register */
 	nxp_set_usb_bits();
 
-	hcd->rsrc_start = pdev->resource[0].start;
-	hcd->rsrc_len = pdev->resource[0].end - pdev->resource[0].start + 1;
-	if (!request_mem_region(hcd->rsrc_start, hcd->rsrc_len, hcd_name)) {
-		dev_dbg(&pdev->dev, "request_mem_region failed\n");
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!res) {
+		dev_err(&pdev->dev, "Failed to get MEM resource");
+		ret =  -ENOMEM;
+		goto out4;
+	}
+
+	hcd->regs = devm_request_and_ioremap(&pdev->dev, res);
+	if (!hcd->regs) {
+		err("Failed to devm_request_and_ioremap");
 		ret =  -ENOMEM;
 		goto out4;
 	}
-	hcd->regs = (void __iomem *)pdev->resource[0].start;
+	hcd->rsrc_start = (u64)(u32)hcd->regs;
+	hcd->rsrc_len = resource_size(res);
 
 	irq = platform_get_irq(pdev, 0);
 	if (irq < 0) {

Re: [PATCH v2 1/8] ohci-nxp: Driver cleanup

[Arnd Bergmann]

On Tuesday 17 April 2012, Roland Stigge wrote:
>Cleanup for ohci-nxp.c:
>
>* Cleanup of resource handling (mem and irq), use devm_request_and_ioremap()
>* Fixed controller initialization (bits)
>* Use DMA_BIT_MASK()
>* Don't remove __devinit
>* Removed obsolete gpio.h include
>* More consistent bit access symbols
>* Removed unused extern declaration ocpi_enable()
>* Added error handling on i2c_get_adapter()
>* Better USB host and device differentiation, since on LPC32xx, the same
>  controller can be used as USB device controller
>* Whitespace cleanup
>
>Signed-off-by: Roland Stigge <stigge@antcom.de>

Looks almost good, I just found one more bug that you introduce:

> +
> +       hcd->regs = devm_request_and_ioremap(&pdev->dev, res);
> +       if (!hcd->regs) {
> +               err("Failed to devm_request_and_ioremap");
>                 ret =  -ENOMEM;
>                 goto out4;
>         }
> -       hcd->regs = (void __iomem *)pdev->resource[0].start;
> +       hcd->rsrc_start = (u64)(u32)hcd->regs;
> +       hcd->rsrc_len = resource_size(res);

This is wrong in multiple ways:

* rsrc_start is a physical address, not an __iomem token.
* you cannot cast a pointer to u32 in general, only to unsigned long.

The fact that you need an ugly type cast like above could have told
you that you are doing something wrong here. I guess it should
be

	hcd->rsrc_start = res->start;

	Arnd

Re: [PATCH v2 1/8] ohci-nxp: Driver cleanup

[Roland Stigge]

Hi,

On 17/04/12 22:36, Arnd Bergmann wrote:
>> +       hcd->regs = devm_request_and_ioremap(&pdev->dev, res);
>> +       if (!hcd->regs) {
>> +               err("Failed to devm_request_and_ioremap");
>>                 ret =  -ENOMEM;
>>                 goto out4;
>>         }
>> -       hcd->regs = (void __iomem *)pdev->resource[0].start;
>> +       hcd->rsrc_start = (u64)(u32)hcd->regs;
>> +       hcd->rsrc_len = resource_size(res);
> 
> This is wrong in multiple ways:
> 
> * rsrc_start is a physical address, not an __iomem token.
> * you cannot cast a pointer to u32 in general, only to unsigned long.
> 
> The fact that you need an ugly type cast like above could have told
> you that you are doing something wrong here. I guess it should
> be
> 
> 	hcd->rsrc_start = res->start;

Right. Thanks for the hint!

Technically, I ported the bug, didn't introduce it ;-)
(pdev->resource[0].start was previously provided via resource already in
IO_ADDRESS (mapped) format). Nevertheless, it's a bug, of course, and it
didn't show up since the driver seems to not use hcd->rsrc_start anyway
(but hcd->regs).

Will post an update after making sure it's tested well.

Roland

Re: [PATCH v2 1/8] ohci-nxp: Driver cleanup

[Dan Carpenter]

On Tue, Apr 17, 2012 at 07:08:20PM +0200, Roland Stigge wrote:
> @@ -435,6 +436,7 @@ static int __devinit usb_hcd_nxp_probe(s
>  	}
>  
>  	__raw_writel(__raw_readl(USB_CTRL) | USB_HOST_NEED_CLK_EN, USB_CTRL);
> +	__raw_writel(__raw_readl(USB_CTRL) & ~USB_DEV_NEED_CLK_EN, USB_CTRL);
>  
>  	/* Set to enable all needed USB clocks */
>  	__raw_writel(USB_CLOCK_MASK, USB_OTG_CLK_CTRL);

What does this chunk do?  Is this part of the "* Fixed controller
initialization (bits)" bullet point?  How does it affect the user?
Should it be included in the 3.4 kernel?

Generally bug fixes, cleanups, and new features should go into
different patches.  If you use "git citool" then you can highlight
and right click on the lines you want to go into the patch.

regards,
dan carpenter

Re: [PATCH v2 1/8] ohci-nxp: Driver cleanup

[Roland Stigge]

On 04/17/2012 10:57 PM, Dan Carpenter wrote:
> On Tue, Apr 17, 2012 at 07:08:20PM +0200, Roland Stigge wrote:
>> @@ -435,6 +436,7 @@ static int __devinit usb_hcd_nxp_probe(s
>>  	}
>>  
>>  	__raw_writel(__raw_readl(USB_CTRL) | USB_HOST_NEED_CLK_EN, USB_CTRL);
>> +	__raw_writel(__raw_readl(USB_CTRL) & ~USB_DEV_NEED_CLK_EN, USB_CTRL);
>>  
>>  	/* Set to enable all needed USB clocks */
>>  	__raw_writel(USB_CLOCK_MASK, USB_OTG_CLK_CTRL);
> 
> What does this chunk do?  Is this part of the "* Fixed controller
> initialization (bits)" bullet point?  How does it affect the user?
> Should it be included in the 3.4 kernel?

Thanks for the hint! I will drop this hunk. Originally, it made sure
that USB_DEV_NEED_CLK_EN is disabled but due to another change:

usb_hcd_nxp_probe() {
...
__raw_writel(USB_SLAVE_HCLK_EN | PAD_CONTROL_LAST_DRIVEN, USB_CTRL);

this bit is already cleared anyway.

(Would have been uncritical for v3.4 anyway, relating only to the other
changes introduced in my 3.5 patches.)

Roland

Re: [PATCH v2 1/8] ohci-nxp: Driver cleanup

[Thierry Reding]

[-- Attachment #1: Type: text/plain, Size: 610 bytes --]

* Roland Stigge wrote:
> --- linux-2.6.orig/drivers/usb/host/ohci-nxp.c
> +++ linux-2.6/drivers/usb/host/ohci-nxp.c
[...]
> @@ -220,7 +220,7 @@ static void isp1301_configure_lpc32xx(vo
>  		ISP1301_I2C_INTERRUPT_RISING | ISP1301_I2C_REG_CLEAR_ADDR, ~0);
>  
>  	/* Enable usb_need_clk clock after transceiver is initialized */
> -	__raw_writel((__raw_readl(USB_CTRL) | (1 << 22)), USB_CTRL);
> +	__raw_writel((__raw_readl(USB_CTRL) | USB_HOST_NEED_CLK_EN), USB_CTRL);

There's a pair of gratuitous parentheses here. Since you're changing the line
anyway you can just drop them as well.

Thierry

[-- Attachment #2: Type: application/pgp-signature, Size: 198 bytes --]

[PATCH v2 2/8] ohci-nxp: Device tree support

[Roland Stigge]

This patch adds device tree support to ohci-nxp.c

Signed-off-by: Roland Stigge <stigge@antcom.de>

---

Applies to v3.4-rc3

Changes since v1:
* Separated out cleanup patch

Thanks to Thierry Reding for reviewing!

 Documentation/devicetree/bindings/usb/ohci-nxp.txt |   16 ++++++++++++++++
 drivers/usb/host/ohci-nxp.c                        |   10 ++++++++++
 2 files changed, 26 insertions(+)

--- /dev/null
+++ linux-2.6/Documentation/devicetree/bindings/usb/ohci-nxp.txt
@@ -0,0 +1,16 @@
+* OHCI controller, NXP ohci-nxp variant
+
+Required properties:
+- compatible: must be "nxp,ohci-nxp"
+- reg: physical base address of the controller and length of memory mapped
+  region.
+- interrupts: The OHCI interrupt
+
+Example (LPC32xx):
+
+	ohci@31020000 {
+		compatible = "nxp,ohci-nxp";
+		reg = <0x31020000 0x300>;
+		interrupt-parent = <&mic>;
+		interrupts = <0x3b 0>;
+	};
--- linux-2.6.orig/drivers/usb/host/ohci-nxp.c
+++ linux-2.6/drivers/usb/host/ohci-nxp.c
@@ -22,6 +22,7 @@
 #include <linux/clk.h>
 #include <linux/platform_device.h>
 #include <linux/i2c.h>
+#include <linux/of.h>
 
 #include <mach/hardware.h>
 #include <asm/mach-types.h>
@@ -526,10 +527,19 @@ static int usb_hcd_nxp_remove(struct pla
 /* work with hotplug and coldplug */
 MODULE_ALIAS("platform:usb-ohci");
 
+#ifdef CONFIG_OF
+static const struct of_device_id usb_hcd_nxp_match[] = {
+	{ .compatible = "nxp,ohci-nxp" },
+	{},
+};
+MODULE_DEVICE_TABLE(of, usb_hcd_nxp_match);
+#endif
+
 static struct platform_driver usb_hcd_nxp_driver = {
 	.driver = {
 		.name = "usb-ohci",
 		.owner	= THIS_MODULE,
+		.of_match_table = of_match_ptr(usb_hcd_nxp_match),
 	},
 	.probe = usb_hcd_nxp_probe,
 	.remove = usb_hcd_nxp_remove,

[PATCH v2 3/8] ARM: LPC32xx: clock.c update

[Roland Stigge]

Besides switching to the default macro CLKDEV_INIT() for clock registration,
this patch actually updates the clock registrations of pl08xdmac.

Signed-off-by: Roland Stigge <stigge@antcom.de>

---

Applies to v3.4-rc3

--- linux-2.6.orig/arch/arm/mach-lpc32xx/clock.c
+++ linux-2.6/arch/arm/mach-lpc32xx/clock.c
@@ -1095,49 +1095,42 @@ struct clk *clk_get_parent(struct clk *c
 }
 EXPORT_SYMBOL(clk_get_parent);
 
-#define _REGISTER_CLOCK(d, n, c) \
-	{ \
-		.dev_id = (d), \
-		.con_id = (n), \
-		.clk = &(c), \
-	},
-
 static struct clk_lookup lookups[] = {
-	_REGISTER_CLOCK(NULL, "osc_32KHz", osc_32KHz)
-	_REGISTER_CLOCK(NULL, "osc_pll397", osc_pll397)
-	_REGISTER_CLOCK(NULL, "osc_main", osc_main)
-	_REGISTER_CLOCK(NULL, "sys_ck", clk_sys)
-	_REGISTER_CLOCK(NULL, "arm_pll_ck", clk_armpll)
-	_REGISTER_CLOCK(NULL, "ck_pll5", clk_usbpll)
-	_REGISTER_CLOCK(NULL, "hclk_ck", clk_hclk)
-	_REGISTER_CLOCK(NULL, "pclk_ck", clk_pclk)
-	_REGISTER_CLOCK(NULL, "timer0_ck", clk_timer0)
-	_REGISTER_CLOCK(NULL, "timer1_ck", clk_timer1)
-	_REGISTER_CLOCK(NULL, "timer2_ck", clk_timer2)
-	_REGISTER_CLOCK(NULL, "timer3_ck", clk_timer3)
-	_REGISTER_CLOCK(NULL, "vfp9_ck", clk_vfp9)
-	_REGISTER_CLOCK(NULL, "clk_dmac", clk_dma)
-	_REGISTER_CLOCK("pnx4008-watchdog", NULL, clk_wdt)
-	_REGISTER_CLOCK(NULL, "uart3_ck", clk_uart3)
-	_REGISTER_CLOCK(NULL, "uart4_ck", clk_uart4)
-	_REGISTER_CLOCK(NULL, "uart5_ck", clk_uart5)
-	_REGISTER_CLOCK(NULL, "uart6_ck", clk_uart6)
-	_REGISTER_CLOCK("pnx-i2c.0", NULL, clk_i2c0)
-	_REGISTER_CLOCK("pnx-i2c.1", NULL, clk_i2c1)
-	_REGISTER_CLOCK("pnx-i2c.2", NULL, clk_i2c2)
-	_REGISTER_CLOCK("dev:ssp0", NULL, clk_ssp0)
-	_REGISTER_CLOCK("dev:ssp1", NULL, clk_ssp1)
-	_REGISTER_CLOCK("lpc32xx_keys.0", NULL, clk_kscan)
-	_REGISTER_CLOCK("lpc32xx-nand.0", "nand_ck", clk_nand)
-	_REGISTER_CLOCK("lpc32xx-adc", NULL, clk_adc)
-	_REGISTER_CLOCK(NULL, "i2s0_ck", clk_i2s0)
-	_REGISTER_CLOCK(NULL, "i2s1_ck", clk_i2s1)
-	_REGISTER_CLOCK("ts-lpc32xx", NULL, clk_tsc)
-	_REGISTER_CLOCK("dev:mmc0", NULL, clk_mmc)
-	_REGISTER_CLOCK("lpc-eth.0", NULL, clk_net)
-	_REGISTER_CLOCK("dev:clcd", NULL, clk_lcd)
-	_REGISTER_CLOCK("lpc32xx_udc", "ck_usbd", clk_usbd)
-	_REGISTER_CLOCK("lpc32xx_rtc", NULL, clk_rtc)
+	CLKDEV_INIT(NULL, "osc_32KHz", &osc_32KHz),
+	CLKDEV_INIT(NULL, "osc_pll397", &osc_pll397),
+	CLKDEV_INIT(NULL, "osc_main", &osc_main),
+	CLKDEV_INIT(NULL, "sys_ck", &clk_sys),
+	CLKDEV_INIT(NULL, "arm_pll_ck", &clk_armpll),
+	CLKDEV_INIT(NULL, "ck_pll5", &clk_usbpll),
+	CLKDEV_INIT(NULL, "hclk_ck", &clk_hclk),
+	CLKDEV_INIT(NULL, "pclk_ck", &clk_pclk),
+	CLKDEV_INIT(NULL, "timer0_ck", &clk_timer0),
+	CLKDEV_INIT(NULL, "timer1_ck", &clk_timer1),
+	CLKDEV_INIT(NULL, "timer2_ck", &clk_timer2),
+	CLKDEV_INIT(NULL, "timer3_ck", &clk_timer3),
+	CLKDEV_INIT(NULL, "vfp9_ck", &clk_vfp9),
+	CLKDEV_INIT("pl08xdmac", NULL, &clk_dma),
+	CLKDEV_INIT("pnx4008-watchdog", NULL, &clk_wdt),
+	CLKDEV_INIT(NULL, "uart3_ck", &clk_uart3),
+	CLKDEV_INIT(NULL, "uart4_ck", &clk_uart4),
+	CLKDEV_INIT(NULL, "uart5_ck", &clk_uart5),
+	CLKDEV_INIT(NULL, "uart6_ck", &clk_uart6),
+	CLKDEV_INIT("pnx-i2c.0", NULL, &clk_i2c0),
+	CLKDEV_INIT("pnx-i2c.1", NULL, &clk_i2c1),
+	CLKDEV_INIT("pnx-i2c.2", NULL, &clk_i2c2),
+	CLKDEV_INIT("dev:ssp0", NULL, &clk_ssp0),
+	CLKDEV_INIT("dev:ssp1", NULL, &clk_ssp1),
+	CLKDEV_INIT("lpc32xx_keys.0", NULL, &clk_kscan),
+	CLKDEV_INIT("lpc32xx-nand.0", "nand_ck", &clk_nand),
+	CLKDEV_INIT("lpc32xx-adc", NULL, &clk_adc),
+	CLKDEV_INIT(NULL, "i2s0_ck", &clk_i2s0),
+	CLKDEV_INIT(NULL, "i2s1_ck", &clk_i2s1),
+	CLKDEV_INIT("ts-lpc32xx", NULL, &clk_tsc),
+	CLKDEV_INIT("dev:mmc0", NULL, &clk_mmc),
+	CLKDEV_INIT("lpc-eth.0", NULL, &clk_net),
+	CLKDEV_INIT("dev:clcd", NULL, &clk_lcd),
+	CLKDEV_INIT("lpc32xx_udc", "ck_usbd", &clk_usbd),
+	CLKDEV_INIT("lpc32xx_rtc", NULL, &clk_rtc),
 };
 
 static int __init clk_init(void)

Re: [PATCH v2 3/8] ARM: LPC32xx: clock.c update

[Dan Carpenter]

On Tue, Apr 17, 2012 at 07:08:22PM +0200, Roland Stigge wrote:
> Besides switching to the default macro CLKDEV_INIT() for clock registration,
> this patch actually updates the clock registrations of pl08xdmac.
> 

When there is a cleanup and a new feature added in the same patch,
it makes it hard to read...  :/  So when we update pl08xdmac:

-       CLKDEV_INIT(NULL, "clk_dmac", &clk_dma),
+       CLKDEV_INIT("pl08xdmac", NULL, &clk_dma),

What does that actually do?  Is it a bug fix?

regards,
dan carpenter

Re: [PATCH v2 3/8] ARM: LPC32xx: clock.c update

[Roland Stigge]

On 04/17/2012 11:09 PM, Dan Carpenter wrote:
> On Tue, Apr 17, 2012 at 07:08:22PM +0200, Roland Stigge wrote:
>> Besides switching to the default macro CLKDEV_INIT() for clock registration,
>> this patch actually updates the clock registrations of pl08xdmac.
>>
> 
> When there is a cleanup and a new feature added in the same patch,
> it makes it hard to read...  :/  So when we update pl08xdmac:
> 
> -       CLKDEV_INIT(NULL, "clk_dmac", &clk_dma),
> +       CLKDEV_INIT("pl08xdmac", NULL, &clk_dma),
> 
> What does that actually do?  Is it a bug fix?

The primecell DMA driver refers to this clock via "pl08xdmac", so it
only works with the above line change. Will separate out to another patch.

Thanks,

Roland

[PATCH v2 4/8] ARM: LPC32xx: Remove obsolete platform Kconfig

[Roland Stigge]

This patch removes two Kconfig options now unused (the network driver is
configured via DT now).

Signed-off-by: Roland Stigge <stigge@antcom.de>

---

Applies to v3.4-rc3

--- linux-2.6.orig/arch/arm/mach-lpc32xx/Kconfig
+++ linux-2.6/arch/arm/mach-lpc32xx/Kconfig
@@ -29,30 +29,4 @@ config ARCH_LPC32XX_UART6_SELECT
 
 endmenu
 
-menu "LPC32XX chip components"
-
-config ARCH_LPC32XX_IRAM_FOR_NET
-	bool "Use IRAM for network buffers"
-	default y
-	help
-	  Say Y here to use the LPC internal fast IRAM (i.e. 256KB SRAM) as
-	  network buffer.  If the total combined required buffer sizes is
-	  larger than the size of IRAM, then SDRAM will be used instead.
-
-	  This can be enabled safely if the IRAM is not intended for other
-	  uses.
-
-config ARCH_LPC32XX_MII_SUPPORT
-	bool "Check to enable MII support or leave disabled for RMII support"
-	help
-	  Say Y here to enable MII support, or N for RMII support. Regardless of
-	  which support is selected, the ethernet interface driver needs to be
-	  selected in the device driver networking section.
-
-	  The PHY3250 reference board uses RMII, so users of this board should
-	  say N.
-
-endmenu
-
 endif
-

[PATCH v2 5/8] input: Device tree support for LPC32xx touchscreen

[Roland Stigge]

This patch adds device tree support for the LPC32xx SoC's touchscreen
controller.

Signed-off-by: Roland Stigge <stigge@antcom.de>
Reviewed-by: Thierry Reding <thierry.reding@avionic-design.de>

---

Applies to v3.4-rc3

 Documentation/devicetree/bindings/input/touchscreen/lpc32xx-tsc.txt |   16 ++++++++++
 drivers/input/touchscreen/lpc32xx_ts.c                              |   10 ++++++
 2 files changed, 26 insertions(+)

--- /dev/null
+++ linux-2.6/Documentation/devicetree/bindings/input/touchscreen/lpc32xx-tsc.txt
@@ -0,0 +1,16 @@
+* NXP LPC32xx SoC Touchscreen Controller (TSC)
+
+Required properties:
+- compatible: must be "nxp,lpc32xx-tsc"
+- reg: physical base address of the controller and length of memory mapped
+  region.
+- interrupts: The TSC/ADC interrupt
+
+Example:
+
+	tsc@40048000 {
+		compatible = "nxp,lpc32xx-tsc";
+		reg = <0x40048000 0x1000>;
+		interrupt-parent = <&mic>;
+		interrupts = <39 0>;
+	};
--- linux-2.6.orig/drivers/input/touchscreen/lpc32xx_ts.c
+++ linux-2.6/drivers/input/touchscreen/lpc32xx_ts.c
@@ -22,6 +22,7 @@
 #include <linux/clk.h>
 #include <linux/io.h>
 #include <linux/slab.h>
+#include <linux/of.h>
 
 /*
  * Touchscreen controller register offsets
@@ -383,6 +384,14 @@ static const struct dev_pm_ops lpc32xx_t
 #define LPC32XX_TS_PM_OPS NULL
 #endif
 
+#ifdef CONFIG_OF
+static struct of_device_id lpc32xx_tsc_of_match[] = {
+	{ .compatible = "nxp,lpc32xx-tsc", },
+	{ },
+};
+MODULE_DEVICE_TABLE(of, lpc32xx_tsc_of_match);
+#endif
+
 static struct platform_driver lpc32xx_ts_driver = {
 	.probe		= lpc32xx_ts_probe,
 	.remove		= __devexit_p(lpc32xx_ts_remove),
@@ -390,6 +399,7 @@ static struct platform_driver lpc32xx_ts
 		.name	= MOD_NAME,
 		.owner	= THIS_MODULE,
 		.pm	= LPC32XX_TS_PM_OPS,
+		.of_match_table = of_match_ptr(lpc32xx_tsc_of_match),
 	},
 };
 module_platform_driver(lpc32xx_ts_driver);

[PATCH v2 6/8] iio: lpc32xx-adc: Remove driver conflict due to device tree

[Roland Stigge]

Previously, the touchscreen and ADC drivers of the LPC32xx SoC had a Kconfig
conflict declared because they use the same hardware. Upon the introduction of
device tree support in both drivers, the conflict must be removed to enable
the same kernel to support different hardware (configured via the device tree).

Signed-off-by: Roland Stigge <stigge@antcom.de>

---

Applies to v3.4-rc3

 drivers/staging/iio/adc/Kconfig |    8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

--- linux-2.6.orig/drivers/staging/iio/adc/Kconfig
+++ linux-2.6/drivers/staging/iio/adc/Kconfig
@@ -195,11 +195,13 @@ config MAX1363_RING_BUFFER
 
 config LPC32XX_ADC
 	tristate "NXP LPC32XX ADC"
-	depends on ARCH_LPC32XX && !TOUCHSCREEN_LPC32XX
+	depends on ARCH_LPC32XX
 	help
 	  Say yes here to build support for the integrated ADC inside the
 	  LPC32XX SoC. Note that this feature uses the same hardware as the
-	  touchscreen driver, so you can only select one of the two drivers
-	  (lpc32xx_adc or lpc32xx_ts). Provides direct access via sysfs.
+	  touchscreen driver, so you should either select only one of the two
+	  drivers (lpc32xx_adc or lpc32xx_ts) or, in the OpenFirmware case,
+	  activate only one via device tree selection.  Provides direct access
+	  via sysfs.
 
 endmenu

[PATCH v2 7/8] ARM: LPC32xx: DTS files for device tree conversion

[Roland Stigge]

This patch adds the dts files for the reference machine of LPC32xx:

* arch/arm/boot/dts/lpc32xx.dtsi: Include for devices based on LPC32xx
* arch/arm/boot/dts/phy3250.dts: Board support for PHYTEC phyCORE-LPC3250

Signed-off-by: Roland Stigge <stigge-uj/7R2tJ6VmzQB+pC5nmwQ@public.gmane.org>

---
 arch/arm/boot/dts/lpc32xx.dtsi |  303 +++++++++++++++++++++++++++++++++++++++++
 arch/arm/boot/dts/phy3250.dts  |  115 +++++++++++++++
 2 files changed, 418 insertions(+)

--- /dev/null
+++ linux-2.6/arch/arm/boot/dts/lpc32xx.dtsi
@@ -0,0 +1,303 @@
+/*
+ * NXP LPC32xx SoC
+ *
+ * Copyright 2012 Roland Stigge <stigge-uj/7R2tJ6VmzQB+pC5nmwQ@public.gmane.org>
+ *
+ * The code contained herein is licensed under the GNU General Public
+ * License. You may obtain a copy of the GNU General Public License
+ * Version 2 or later at the following locations:
+ *
+ * http://www.opensource.org/licenses/gpl-license.html
+ * http://www.gnu.org/copyleft/gpl.html
+ */
+
+/include/ "skeleton.dtsi"
+
+/ {
+	compatible = "nxp,lpc32xx";
+
+	cpus {
+		cpu@0 {
+			compatible = "arm,arm926ejs";
+		};
+	};
+
+	ahb {
+		#address-cells = <1>;
+		#size-cells = <1>;
+		compatible = "simple-bus";
+		ranges = <0x20000000 0x20000000 0x30000000>;
+
+		slc: flash@20020000 {
+			compatible = "nxp,nand-slc";
+			reg = <0x20020000 0x1000>;
+			status = "disable";
+		};
+
+		mlc: flash@200B0000 {
+			compatible = "nxp,nand-mlc";
+			reg = <0x200B0000 0x1000>;
+			status = "disable";
+		};
+
+		dma@31000000 {
+			compatible = "arm,pl080", "arm,primecell";
+			reg = <0x31000000 0x1000>;
+			interrupt-parent = <&mic>;
+			interrupts = <0x1c 0>;
+		};
+
+		/*
+		 * Enable either ohci or usbd (gadget)!
+		 */
+		ohci@31020000 {
+			compatible = "nxp,ohci-nxp", "usb-ohci";
+			reg = <0x31020000 0x300>;
+			interrupt-parent = <&mic>;
+			interrupts = <0x3b 0>;
+			status = "disable";
+		};
+
+		usbd@31020000 {
+			compatible = "nxp,lpc32xx-udc";
+			reg = <0x31020000 0x300>;
+			interrupt-parent = <&mic>;
+			interrupts = <0x3d 0>, <0x3e 0>, <0x3c 0>, <0x3a 0>;
+			status = "disable";
+		};
+
+		clcd@31040000 {
+			compatible = "arm,pl110", "arm,primecell";
+			reg = <0x31040000 0x1000>;
+			interrupt-parent = <&mic>;
+			interrupts = <0x0e 0>;
+			status = "disable";
+		};
+
+		mac: ethernet@31060000 {
+			compatible = "nxp,lpc-eth";
+			reg = <0x31060000 0x1000>;
+			interrupt-parent = <&mic>;
+			interrupts = <0x1d 0>;
+		};
+
+		apb {
+			#address-cells = <1>;
+			#size-cells = <1>;
+			compatible = "simple-bus";
+			ranges = <0x20000000 0x20000000 0x30000000>;
+
+			ssp0: ssp@20084000 {
+				compatible = "arm,pl022", "arm,primecell";
+				reg = <0x20084000 0x1000>;
+				interrupt-parent = <&mic>;
+				interrupts = <0x14 0>;
+			};
+
+			spi1: spi@20088000 {
+				compatible = "nxp,spi";
+				reg = <0x20088000 0x1000>;
+			};
+
+			ssp1: ssp@2008c000 {
+				compatible = "arm,pl022", "arm,primecell";
+				reg = <0x2008c000 0x1000>;
+				interrupt-parent = <&mic>;
+				interrupts = <0x15 0>;
+			};
+
+			spi2: spi@20090000 {
+				compatible = "nxp,spi";
+				reg = <0x20090000 0x1000>;
+			};
+
+			i2s0: i2s@20094000 {
+				compatible = "nxp,i2s";
+				reg = <0x20094000 0x1000>;
+			};
+
+			sd@20098000 {
+				compatible = "arm,pl180", "arm,primecell";
+				reg = <0x20098000 0x1000>;
+				interrupt-parent = <&mic>;
+				interrupts = <0x0f 0>, <0x0d 0>;
+			};
+
+			i2s1: i2s@2009C000 {
+				compatible = "nxp,i2s";
+				reg = <0x2009C000 0x1000>;
+			};
+
+			uart3: serial@40080000 {
+				compatible = "nxp,serial";
+				reg = <0x40080000 0x1000>;
+			};
+
+			uart4: serial@40088000 {
+				compatible = "nxp,serial";
+				reg = <0x40088000 0x1000>;
+			};
+
+			uart5: serial@40090000 {
+				compatible = "nxp,serial";
+				reg = <0x40090000 0x1000>;
+			};
+
+			uart6: serial@40098000 {
+				compatible = "nxp,serial";
+				reg = <0x40098000 0x1000>;
+			};
+
+			i2c1: i2c@400A0000 {
+				compatible = "nxp,pnx-i2c";
+				reg = <0x400A0000 0x100>;
+				interrupt-parent = <&mic>;
+				interrupts = <0x33 0>;
+				#address-cells = <1>;
+				#size-cells = <0>;
+				pnx,timeout = <0x64>;
+			};
+
+			i2c2: i2c@400A8000 {
+				compatible = "nxp,pnx-i2c";
+				reg = <0x400A8000 0x100>;
+				interrupt-parent = <&mic>;
+				interrupts = <0x32 0>;
+				#address-cells = <1>;
+				#size-cells = <0>;
+				pnx,timeout = <0x64>;
+			};
+
+			i2cusb: i2c@31020300 {
+				compatible = "nxp,pnx-i2c";
+				reg = <0x31020300 0x100>;
+				interrupt-parent = <&mic>;
+				interrupts = <0x3f 0>;
+				#address-cells = <1>;
+				#size-cells = <0>;
+				pnx,timeout = <0x64>;
+			};
+		};
+
+		fab {
+			#address-cells = <1>;
+			#size-cells = <1>;
+			compatible = "simple-bus";
+			ranges = <0x20000000 0x20000000 0x30000000>;
+
+			/*
+			 * MIC Interrupt controller includes:
+			 *   MIC @40008000
+			 *   SIC1 @4000C000
+			 *   SIC2 @40010000
+			 */
+			mic: interrupt-controller@40008000 {
+				compatible = "nxp,lpc32xx-mic";
+				interrupt-controller;
+				interrupt-parent;
+				reg = <0x40008000 0xC000>;
+				#interrupt-cells = <2>;
+			};
+
+			uart1: serial@40014000 {
+				compatible = "nxp,serial";
+				reg = <0x40014000 0x1000>;
+			};
+
+			uart2: serial@40018000 {
+				compatible = "nxp,serial";
+				reg = <0x40018000 0x1000>;
+			};
+
+			uart7: serial@4001C000 {
+				compatible = "nxp,serial";
+				reg = <0x4001C000 0x1000>;
+			};
+
+			rtc@40024000 {
+				compatible = "nxp,lpc32xx-rtc";
+				reg = <0x40024000 0x1000>;
+				interrupt-parent = <&mic>;
+				interrupts = <0x34 0>;
+			};
+
+			gpio: gpio@40028000 {
+				compatible = "nxp,lpc32xx-gpio";
+				reg = <0x40028000 0x1000>;
+				/* create a private address space for enumeration */
+				#address-cells = <1>;
+				#size-cells = <0>;
+
+				gpio_p0: gpio-bank@0 {
+					gpio-controller;
+					#gpio-cells = <2>;
+					reg = <0>;
+				};
+
+				gpio_p1: gpio-bank@1 {
+					gpio-controller;
+					#gpio-cells = <2>;
+					reg = <1>;
+				};
+
+				gpio_p2: gpio-bank@2 {
+					gpio-controller;
+					#gpio-cells = <2>;
+					reg = <2>;
+				};
+
+				gpio_p3: gpio-bank@3 {
+					gpio-controller;
+					#gpio-cells = <2>;
+					reg = <3>;
+				};
+
+				gpi_p3: gpio-bank@4 {
+					gpio-controller;
+					#gpio-cells = <2>;
+					reg = <4>;
+				};
+
+				gpo_p3: gpio-bank@5 {
+					gpio-controller;
+					#gpio-cells = <2>;
+					reg = <5>;
+				};
+			};
+
+			watchdog@4003C000 {
+				compatible = "nxp,pnx4008-wdt";
+				reg = <0x4003C000 0x1000>;
+			};
+
+			/*
+			 * TSC vs. ADC: Since those two share the same
+			 * hardware, you need to choose from one of the
+			 * following two and do 'status = "okay";" for one of
+			 * them
+			 */
+
+			adc@40048000 {
+				compatible = "nxp,lpc32xx-adc";
+				reg = <0x40048000 0x1000>;
+				interrupt-parent = <&mic>;
+				interrupts = <0x27 0>;
+				status = "disable";
+			};
+
+			tsc@40048000 {
+				compatible = "nxp,lpc32xx-tsc";
+				reg = <0x40048000 0x1000>;
+				interrupt-parent = <&mic>;
+				interrupts = <0x27 0>;
+				status = "disable";
+			};
+
+			key@40050000 {
+				compatible = "nxp,key";
+				reg = <0x40050000 0x1000>;
+			};
+
+		};
+	};
+};
--- /dev/null
+++ linux-2.6/arch/arm/boot/dts/phy3250.dts
@@ -0,0 +1,115 @@
+/*
+ * PHYTEC phyCORE-LPC3250 board
+ *
+ * Copyright 2012 Roland Stigge <stigge-uj/7R2tJ6VmzQB+pC5nmwQ@public.gmane.org>
+ *
+ * The code contained herein is licensed under the GNU General Public
+ * License. You may obtain a copy of the GNU General Public License
+ * Version 2 or later at the following locations:
+ *
+ * http://www.opensource.org/licenses/gpl-license.html
+ * http://www.gnu.org/copyleft/gpl.html
+ */
+
+/dts-v1/;
+/include/ "lpc32xx.dtsi"
+
+/ {
+	model = "PHYTEC phyCORE-LPC3250 board based on NXP LPC3250";
+	compatible = "phytec,phy3250", "nxp,lpc32xx";
+	#address-cells = <1>;
+	#size-cells = <1>;
+
+	memory {
+		device_type = "memory";
+		reg = <0 0x4000000>;
+	};
+
+	ahb {
+		mac: ethernet@31060000 {
+			phy-mode = "rmii";
+			use-iram;
+		};
+
+		/* Here, choose from: ohci, usbd */
+		ohci@31020000 {
+			transceiver = <&isp1301>;
+			status = "okay";
+		};
+
+/*
+		usbd@31020000 {
+			transceiver = <&isp1301>;
+			status = "okay";
+		};
+*/
+
+		clcd@31040000 {
+			status = "okay";
+		};
+
+		apb {
+			i2c1: i2c@400A0000 {
+				clock-frequency = <0x186a0>;
+
+				pcf8563: rtc@51 {
+					compatible = "nxp,pcf8563";
+					reg = <0x51>;
+				};
+
+				uda1380: uda1380@18 {
+					compatible = "nxp,uda1380";
+					reg = <0x18>;
+					power-gpio = <&gpio 0x59 0>;
+					reset-gpio = <&gpio 0x51 0>;
+					dac-clk = "wspll";
+				};
+			};
+
+			i2c2: i2c@400A8000 {
+				clock-frequency = <0x186a0>;
+			};
+
+			i2cusb: i2c@31020300 {
+				clock-frequency = <0x186a0>;
+
+				isp1301: usb-transceiver@2c {
+					compatible = "nxp,isp1301";
+					reg = <0x2c>;
+				};
+			};
+
+			ssp0: ssp@20084000 {
+				eeprom: at25@0 {
+					compatible = "atmel,at25";
+				};
+			};
+		};
+
+		fab {
+			tsc@40048000 {
+				status = "okay";
+			};
+		};
+	};
+
+	leds {
+		compatible = "gpio-leds";
+
+		led0 {
+			gpios = <&gpo_p3 1 1>; /* GPO_P3 1, GPIO 80, active low */
+			linux,default-trigger = "heartbeat";
+			default-state = "off";
+		};
+
+		led1 {
+			gpios = <&gpo_p3 14 1>; /* GPO_P3 14, GPIO 93, active low */
+			linux,default-trigger = "timer";
+			default-state = "off";
+		};
+	};
+
+	chosen {
+		bootargs = "console=ttyS0,115200n81 root=/dev/ram0";
+	};
+};
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v2 7/8] ARM: LPC32xx: DTS files for device tree conversion

[Thierry Reding]

[-- Attachment #1: Type: text/plain, Size: 1874 bytes --]

* Roland Stigge wrote:
[...]
> --- /dev/null
> +++ linux-2.6/arch/arm/boot/dts/lpc32xx.dtsi
> @@ -0,0 +1,303 @@
> +/*
> + * NXP LPC32xx SoC
> + *
> + * Copyright 2012 Roland Stigge <stigge-uj/7R2tJ6VmzQB+pC5nmwQ@public.gmane.org>
> + *
> + * The code contained herein is licensed under the GNU General Public
> + * License. You may obtain a copy of the GNU General Public License
> + * Version 2 or later at the following locations:
> + *
> + * http://www.opensource.org/licenses/gpl-license.html
> + * http://www.gnu.org/copyleft/gpl.html
> + */
> +
> +/include/ "skeleton.dtsi"
> +
> +/ {
> +	compatible = "nxp,lpc32xx";
> +
> +	cpus {
> +		cpu@0 {
> +			compatible = "arm,arm926ejs";
> +		};
> +	};
> +
> +	ahb {
> +		#address-cells = <1>;
> +		#size-cells = <1>;
> +		compatible = "simple-bus";
> +		ranges = <0x20000000 0x20000000 0x30000000>;
> +
> +		slc: flash@20020000 {
> +			compatible = "nxp,nand-slc";
> +			reg = <0x20020000 0x1000>;
> +			status = "disable";
> +		};
> +
> +		mlc: flash@200B0000 {
> +			compatible = "nxp,nand-mlc";
> +			reg = <0x200B0000 0x1000>;
> +			status = "disable";
> +		};
> +
> +		dma@31000000 {
> +			compatible = "arm,pl080", "arm,primecell";
> +			reg = <0x31000000 0x1000>;
> +			interrupt-parent = <&mic>;

Most platforms define the interrupt parent somewhere in the top-level node
and have it inherited for child devices instead of having to repeat it for
each device. But I guess it's also okay to make it explicit.

[...]
> +			/*
> +			 * MIC Interrupt controller includes:
> +			 *   MIC @40008000
> +			 *   SIC1 @4000C000
> +			 *   SIC2 @40010000
> +			 */
> +			mic: interrupt-controller@40008000 {
> +				compatible = "nxp,lpc32xx-mic";
> +				interrupt-controller;
> +				interrupt-parent;

I /think/ in this case you can leave this out. At least there doesn't seem to
be any other such usage in the tree.

Thierry

[-- Attachment #2: Type: application/pgp-signature, Size: 198 bytes --]

Re: [PATCH v2 7/8] ARM: LPC32xx: DTS files for device tree conversion

[Arnd Bergmann]

On Wednesday 18 April 2012, Thierry Reding wrote:
> [...]
> > +                     /*
> > +                      * MIC Interrupt controller includes:
> > +                      *   MIC @40008000
> > +                      *   SIC1 @4000C000
> > +                      *   SIC2 @40010000
> > +                      */
> > +                     mic: interrupt-controller@40008000 {
> > +                             compatible = "nxp,lpc32xx-mic";
> > +                             interrupt-controller;
> > +                             interrupt-parent;
> 
> I think in this case you can leave this out. At least there doesn't seem to
> be any other such usage in the tree.
> 

It's actually a bug and we're removing all empty "interrupt-parent"
properties from the dts files now because they cause problems. Please don't
introduce new ones.

	Arnd

--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v2 7/8] ARM: LPC32xx: DTS files for device tree conversion

[Arnd Bergmann]

On Tuesday 17 April 2012, Roland Stigge wrote:
> +	ahb {
> +		#address-cells = <1>;
> +		#size-cells = <1>;
> +		compatible = "simple-bus";
> +		ranges = <0x20000000 0x20000000 0x30000000>;
> +
> +		slc: flash@20020000 {
> +			compatible = "nxp,nand-slc";
> +			reg = <0x20020000 0x1000>;
> +			status = "disable";
> +		};
> +
> +		mlc: flash@200B0000 {
> +			compatible = "nxp,nand-mlc";
> +			reg = <0x200B0000 0x1000>;
> +			status = "disable";
> +		};

You have the flash chips in here, which seem right, but I would expect
to see the phy3250 file pick this up again and define partitions in
it, at least if there is a well-defined partition layout for that
board.

> +
> +	chosen {
> +		bootargs = "console=ttyS0,115200n81 root=/dev/ram0";
> +	};

I think in general it's better not to list a command line like this,
it should instead get set by the boot loader or using ARM_ATAG_DTB_COMPAT.

	Arnd

[PATCH v2 8/8] ARM: LPC32xx: Device tree support

[Roland Stigge]

This patch does the actual device tree switch for the LPC32xx SoC.

Signed-off-by: Roland Stigge <stigge@antcom.de>

---

Applies to v3.4-rc3

 Documentation/devicetree/bindings/arm/lpc32xx-mic.txt |   38 +++
 Documentation/devicetree/bindings/arm/lpc32xx.txt     |    8 
 arch/arm/Kconfig                                      |    1 
 arch/arm/mach-lpc32xx/Kconfig                         |   26 --
 arch/arm/mach-lpc32xx/clock.c                         |   77 +++----
 arch/arm/mach-lpc32xx/common.c                        |  192 ------------------
 arch/arm/mach-lpc32xx/common.h                        |   14 -
 arch/arm/mach-lpc32xx/irq.c                           |   78 +++++--
 arch/arm/mach-lpc32xx/phy3250.c                       |  146 +++++--------
 9 files changed, 202 insertions(+), 378 deletions(-)

--- /dev/null
+++ linux-2.6/Documentation/devicetree/bindings/arm/lpc32xx-mic.txt
@@ -0,0 +1,38 @@
+* NXP LPC32xx Main Interrupt Controller
+  (MIC, including SIC1 and SIC2 secondary controllers)
+
+Required properties:
+- compatible: Should be "nxp,lpc32xx-mic"
+- interrupt-controller: Identifies the node as an interrupt controller.
+- interrupt-parent: Empty for the interrupt controller itself
+- #interrupt-cells: The number of cells to define the interrupts. Should be 2.
+  The first cell is the IRQ number
+  The second cell is used to specify mode:
+      1 = low-to-high edge triggered
+      2 = high-to-low edge triggered
+      4 = active high level-sensitive
+      8 = active low level-sensitive
+      Default for internal sources should be set to 4 (active high).
+- reg: Should contain MIC registers location and length
+
+Examples:
+	/*
+	 * MIC
+	 */
+	mic: interrupt-controller@40008000 {
+		compatible = "nxp,lpc32xx-mic";
+		interrupt-controller;
+		interrupt-parent;
+		#interrupt-cells = <2>;
+		reg = <0x40008000 0xC000>;
+	};
+
+	/*
+	 * ADC
+	 */
+	adc@40048000 {
+		compatible = "nxp,lpc32xx-adc";
+		reg = <0x40048000 0x1000>;
+		interrupt-parent = <&mic>;
+		interrupts = <39 4>;
+	};
--- /dev/null
+++ linux-2.6/Documentation/devicetree/bindings/arm/lpc32xx.txt
@@ -0,0 +1,8 @@
+NXP LPC32xx Platforms Device Tree Bindings
+------------------------------------------
+
+Boards with the NXP LPC32xx SoC shall have the following properties:
+
+Required root node property:
+
+compatible = "nxp,lpc32xx";
--- linux-2.6.orig/arch/arm/Kconfig
+++ linux-2.6/arch/arm/Kconfig
@@ -597,6 +597,7 @@ config ARCH_LPC32XX
 	select USB_ARCH_HAS_OHCI
 	select CLKDEV_LOOKUP
 	select GENERIC_CLOCKEVENTS
+	select USE_OF
 	help
 	  Support for the NXP LPC32XX family of processors
 
--- linux-2.6.orig/arch/arm/mach-lpc32xx/common.c
+++ linux-2.6/arch/arm/mach-lpc32xx/common.c
@@ -32,198 +32,6 @@
 #include "common.h"
 
 /*
- * Watchdog timer
- */
-static struct resource watchdog_resources[] = {
-	[0] = {
-		.start = LPC32XX_WDTIM_BASE,
-		.end = LPC32XX_WDTIM_BASE + SZ_4K - 1,
-		.flags = IORESOURCE_MEM,
-	},
-};
-
-struct platform_device lpc32xx_watchdog_device = {
-	.name = "pnx4008-watchdog",
-	.id = -1,
-	.num_resources = ARRAY_SIZE(watchdog_resources),
-	.resource = watchdog_resources,
-};
-
-/*
- * I2C busses
- */
-static struct resource i2c0_resources[] = {
-	[0] = {
-		.start = LPC32XX_I2C1_BASE,
-		.end = LPC32XX_I2C1_BASE + 0x100 - 1,
-		.flags = IORESOURCE_MEM,
-	},
-	[1] = {
-		.start = IRQ_LPC32XX_I2C_1,
-		.end = IRQ_LPC32XX_I2C_1,
-		.flags = IORESOURCE_IRQ,
-	},
-};
-
-static struct resource i2c1_resources[] = {
-	[0] = {
-		.start = LPC32XX_I2C2_BASE,
-		.end = LPC32XX_I2C2_BASE + 0x100 - 1,
-		.flags = IORESOURCE_MEM,
-	},
-	[1] = {
-		.start = IRQ_LPC32XX_I2C_2,
-		.end = IRQ_LPC32XX_I2C_2,
-		.flags = IORESOURCE_IRQ,
-	},
-};
-
-static struct resource i2c2_resources[] = {
-	[0] = {
-		.start = LPC32XX_OTG_I2C_BASE,
-		.end = LPC32XX_OTG_I2C_BASE + 0x100 - 1,
-		.flags = IORESOURCE_MEM,
-	},
-	[1] = {
-		.start = IRQ_LPC32XX_USB_I2C,
-		.end = IRQ_LPC32XX_USB_I2C,
-		.flags = IORESOURCE_IRQ,
-	},
-};
-
-struct platform_device lpc32xx_i2c0_device = {
-	.name = "pnx-i2c",
-	.id = 0,
-	.num_resources = ARRAY_SIZE(i2c0_resources),
-	.resource = i2c0_resources,
-};
-
-struct platform_device lpc32xx_i2c1_device = {
-	.name = "pnx-i2c",
-	.id = 1,
-	.num_resources = ARRAY_SIZE(i2c1_resources),
-	.resource = i2c1_resources,
-};
-
-struct platform_device lpc32xx_i2c2_device = {
-	.name = "pnx-i2c",
-	.id = 2,
-	.num_resources = ARRAY_SIZE(i2c2_resources),
-	.resource = i2c2_resources,
-};
-
-/* TSC (Touch Screen Controller) */
-
-static struct resource lpc32xx_tsc_resources[] = {
-	{
-		.start = LPC32XX_ADC_BASE,
-		.end = LPC32XX_ADC_BASE + SZ_4K - 1,
-		.flags = IORESOURCE_MEM,
-	}, {
-		.start = IRQ_LPC32XX_TS_IRQ,
-		.end = IRQ_LPC32XX_TS_IRQ,
-		.flags = IORESOURCE_IRQ,
-	},
-};
-
-struct platform_device lpc32xx_tsc_device = {
-	.name =  "ts-lpc32xx",
-	.id = -1,
-	.num_resources = ARRAY_SIZE(lpc32xx_tsc_resources),
-	.resource = lpc32xx_tsc_resources,
-};
-
-/* RTC */
-
-static struct resource lpc32xx_rtc_resources[] = {
-	{
-		.start = LPC32XX_RTC_BASE,
-		.end = LPC32XX_RTC_BASE + SZ_4K - 1,
-		.flags = IORESOURCE_MEM,
-	},{
-		.start = IRQ_LPC32XX_RTC,
-		.end = IRQ_LPC32XX_RTC,
-		.flags = IORESOURCE_IRQ,
-	},
-};
-
-struct platform_device lpc32xx_rtc_device = {
-	.name =  "rtc-lpc32xx",
-	.id = -1,
-	.num_resources = ARRAY_SIZE(lpc32xx_rtc_resources),
-	.resource = lpc32xx_rtc_resources,
-};
-
-/*
- * ADC support
- */
-static struct resource adc_resources[] = {
-	{
-		.start = LPC32XX_ADC_BASE,
-		.end = LPC32XX_ADC_BASE + SZ_4K - 1,
-		.flags = IORESOURCE_MEM,
-	}, {
-		.start = IRQ_LPC32XX_TS_IRQ,
-		.end = IRQ_LPC32XX_TS_IRQ,
-		.flags = IORESOURCE_IRQ,
-	},
-};
-
-struct platform_device lpc32xx_adc_device = {
-	.name =  "lpc32xx-adc",
-	.id = -1,
-	.num_resources = ARRAY_SIZE(adc_resources),
-	.resource = adc_resources,
-};
-
-/*
- * USB support
- */
-/* The dmamask must be set for OHCI to work */
-static u64 ohci_dmamask = ~(u32) 0;
-static struct resource ohci_resources[] = {
-	{
-		.start = IO_ADDRESS(LPC32XX_USB_BASE),
-		.end = IO_ADDRESS(LPC32XX_USB_BASE + 0x100 - 1),
-		.flags = IORESOURCE_MEM,
-	}, {
-		.start = IRQ_LPC32XX_USB_HOST,
-		.flags = IORESOURCE_IRQ,
-	},
-};
-struct platform_device lpc32xx_ohci_device = {
-	.name = "usb-ohci",
-	.id = -1,
-	.dev = {
-		.dma_mask = &ohci_dmamask,
-		.coherent_dma_mask = 0xFFFFFFFF,
-	},
-	.num_resources = ARRAY_SIZE(ohci_resources),
-	.resource = ohci_resources,
-};
-
-/*
- * Network Support
- */
-static struct resource net_resources[] = {
-	[0] = DEFINE_RES_MEM(LPC32XX_ETHERNET_BASE, SZ_4K),
-	[1] = DEFINE_RES_MEM(LPC32XX_IRAM_BASE, SZ_128K),
-	[2] = DEFINE_RES_IRQ(IRQ_LPC32XX_ETHERNET),
-};
-
-static u64 lpc32xx_mac_dma_mask = 0xffffffffUL;
-struct platform_device lpc32xx_net_device = {
-	.name = "lpc-eth",
-	.id = 0,
-	.dev = {
-		.dma_mask = &lpc32xx_mac_dma_mask,
-		.coherent_dma_mask = 0xffffffffUL,
-	},
-	.num_resources = ARRAY_SIZE(net_resources),
-	.resource = net_resources,
-};
-
-/*
  * Returns the unique ID for the device
  */
 void lpc32xx_get_uid(u32 devid[4])
--- linux-2.6.orig/arch/arm/mach-lpc32xx/common.h
+++ linux-2.6/arch/arm/mach-lpc32xx/common.h
@@ -23,26 +23,12 @@
 #include <linux/platform_device.h>
 
 /*
- * Arch specific platform device structures
- */
-extern struct platform_device lpc32xx_watchdog_device;
-extern struct platform_device lpc32xx_i2c0_device;
-extern struct platform_device lpc32xx_i2c1_device;
-extern struct platform_device lpc32xx_i2c2_device;
-extern struct platform_device lpc32xx_tsc_device;
-extern struct platform_device lpc32xx_adc_device;
-extern struct platform_device lpc32xx_rtc_device;
-extern struct platform_device lpc32xx_ohci_device;
-extern struct platform_device lpc32xx_net_device;
-
-/*
  * Other arch specific structures and functions
  */
 extern struct sys_timer lpc32xx_timer;
 extern void __init lpc32xx_init_irq(void);
 extern void __init lpc32xx_map_io(void);
 extern void __init lpc32xx_serial_init(void);
-extern void __init lpc32xx_gpio_init(void);
 extern void lpc23xx_restart(char, const char *);
 
 
--- linux-2.6.orig/arch/arm/mach-lpc32xx/irq.c
+++ linux-2.6/arch/arm/mach-lpc32xx/irq.c
@@ -22,6 +22,11 @@
 #include <linux/irq.h>
 #include <linux/err.h>
 #include <linux/io.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/of_irq.h>
+#include <linux/irqdomain.h>
+#include <linux/module.h>
 
 #include <mach/irqs.h>
 #include <mach/hardware.h>
@@ -44,6 +49,9 @@
 #define SIC1_ATR_DEFAULT	0x00026000
 #define SIC2_ATR_DEFAULT	0x00000000
 
+static struct irq_domain *lpc32xx_mic_domain;
+static struct device_node *lpc32xx_mic_np;
+
 struct lpc32xx_event_group_regs {
 	void __iomem *enab_reg;
 	void __iomem *edge_reg;
@@ -203,7 +211,7 @@ static void lpc32xx_mask_irq(struct irq_
 {
 	unsigned int reg, ctrl, mask;
 
-	get_controller(d->irq, &ctrl, &mask);
+	get_controller(d->hwirq, &ctrl, &mask);
 
 	reg = __raw_readl(LPC32XX_INTC_MASK(ctrl)) & ~mask;
 	__raw_writel(reg, LPC32XX_INTC_MASK(ctrl));
@@ -213,7 +221,7 @@ static void lpc32xx_unmask_irq(struct ir
 {
 	unsigned int reg, ctrl, mask;
 
-	get_controller(d->irq, &ctrl, &mask);
+	get_controller(d->hwirq, &ctrl, &mask);
 
 	reg = __raw_readl(LPC32XX_INTC_MASK(ctrl)) | mask;
 	__raw_writel(reg, LPC32XX_INTC_MASK(ctrl));
@@ -223,14 +231,14 @@ static void lpc32xx_ack_irq(struct irq_d
 {
 	unsigned int ctrl, mask;
 
-	get_controller(d->irq, &ctrl, &mask);
+	get_controller(d->hwirq, &ctrl, &mask);
 
 	__raw_writel(mask, LPC32XX_INTC_RAW_STAT(ctrl));
 
 	/* Also need to clear pending wake event */
-	if (lpc32xx_events[d->irq].mask != 0)
-		__raw_writel(lpc32xx_events[d->irq].mask,
-			lpc32xx_events[d->irq].event_group->rawstat_reg);
+	if (lpc32xx_events[d->hwirq].mask != 0)
+		__raw_writel(lpc32xx_events[d->hwirq].mask,
+			lpc32xx_events[d->hwirq].event_group->rawstat_reg);
 }
 
 static void __lpc32xx_set_irq_type(unsigned int irq, int use_high_level,
@@ -274,22 +282,22 @@ static int lpc32xx_set_irq_type(struct i
 	switch (type) {
 	case IRQ_TYPE_EDGE_RISING:
 		/* Rising edge sensitive */
-		__lpc32xx_set_irq_type(d->irq, 1, 1);
+		__lpc32xx_set_irq_type(d->hwirq, 1, 1);
 		break;
 
 	case IRQ_TYPE_EDGE_FALLING:
 		/* Falling edge sensitive */
-		__lpc32xx_set_irq_type(d->irq, 0, 1);
+		__lpc32xx_set_irq_type(d->hwirq, 0, 1);
 		break;
 
 	case IRQ_TYPE_LEVEL_LOW:
 		/* Low level sensitive */
-		__lpc32xx_set_irq_type(d->irq, 0, 0);
+		__lpc32xx_set_irq_type(d->hwirq, 0, 0);
 		break;
 
 	case IRQ_TYPE_LEVEL_HIGH:
 		/* High level sensitive */
-		__lpc32xx_set_irq_type(d->irq, 1, 0);
+		__lpc32xx_set_irq_type(d->hwirq, 1, 0);
 		break;
 
 	/* Other modes are not supported */
@@ -298,7 +306,7 @@ static int lpc32xx_set_irq_type(struct i
 	}
 
 	/* Ok to use the level handler for all types */
-	irq_set_handler(d->irq, handle_level_irq);
+	irq_set_handler(d->hwirq, handle_level_irq);
 
 	return 0;
 }
@@ -307,33 +315,33 @@ static int lpc32xx_irq_wake(struct irq_d
 {
 	unsigned long eventreg;
 
-	if (lpc32xx_events[d->irq].mask != 0) {
-		eventreg = __raw_readl(lpc32xx_events[d->irq].
+	if (lpc32xx_events[d->hwirq].mask != 0) {
+		eventreg = __raw_readl(lpc32xx_events[d->hwirq].
 			event_group->enab_reg);
 
 		if (state)
-			eventreg |= lpc32xx_events[d->irq].mask;
+			eventreg |= lpc32xx_events[d->hwirq].mask;
 		else {
-			eventreg &= ~lpc32xx_events[d->irq].mask;
+			eventreg &= ~lpc32xx_events[d->hwirq].mask;
 
 			/*
 			 * When disabling the wakeup, clear the latched
 			 * event
 			 */
-			__raw_writel(lpc32xx_events[d->irq].mask,
-				lpc32xx_events[d->irq].
+			__raw_writel(lpc32xx_events[d->hwirq].mask,
+				lpc32xx_events[d->hwirq].
 				event_group->rawstat_reg);
 		}
 
 		__raw_writel(eventreg,
-			lpc32xx_events[d->irq].event_group->enab_reg);
+			lpc32xx_events[d->hwirq].event_group->enab_reg);
 
 		return 0;
 	}
 
 	/* Clear event */
-	__raw_writel(lpc32xx_events[d->irq].mask,
-		lpc32xx_events[d->irq].event_group->rawstat_reg);
+	__raw_writel(lpc32xx_events[d->hwirq].mask,
+		lpc32xx_events[d->hwirq].event_group->rawstat_reg);
 
 	return -ENODEV;
 }
@@ -353,6 +361,7 @@ static void __init lpc32xx_set_default_m
 }
 
 static struct irq_chip lpc32xx_irq_chip = {
+	.name = "MIC",
 	.irq_ack = lpc32xx_ack_irq,
 	.irq_mask = lpc32xx_mask_irq,
 	.irq_unmask = lpc32xx_unmask_irq,
@@ -386,9 +395,23 @@ static void lpc32xx_sic2_handler(unsigne
 	}
 }
 
+static int __init __lpc32xx_mic_of_init(struct device_node *node,
+					struct device_node *parent)
+{
+	lpc32xx_mic_np = node;
+
+	return 0;
+}
+
+static const struct of_device_id mic_of_match[] __initconst = {
+	{ .compatible = "nxp,lpc32xx-mic", .data = __lpc32xx_mic_of_init },
+	{ }
+};
+
 void __init lpc32xx_init_irq(void)
 {
 	unsigned int i;
+	int irq_base;
 
 	/* Setup MIC */
 	__raw_writel(0, LPC32XX_INTC_MASK(LPC32XX_MIC_BASE));
@@ -448,4 +471,19 @@ void __init lpc32xx_init_irq(void)
 		LPC32XX_CLKPWR_PIN_RS);
 	__raw_writel(__raw_readl(LPC32XX_CLKPWR_INT_RS),
 		LPC32XX_CLKPWR_INT_RS);
+
+	of_irq_init(mic_of_match);
+
+	irq_base = irq_alloc_descs(-1, 0, NR_IRQS, 0);
+	if (irq_base < 0) {
+		pr_warn("Cannot allocate irq_descs, assuming pre-allocated\n");
+		irq_base = 0;
+	}
+
+	lpc32xx_mic_domain = irq_domain_add_legacy(lpc32xx_mic_np, NR_IRQS,
+						   irq_base, 0,
+						   &irq_domain_simple_ops,
+						   NULL);
+	if (!lpc32xx_mic_domain)
+		panic("Unable to add MIC irq domain\n");
 }
--- linux-2.6.orig/arch/arm/mach-lpc32xx/phy3250.c
+++ linux-2.6/arch/arm/mach-lpc32xx/phy3250.c
@@ -1,8 +1,9 @@
 /*
- * arch/arm/mach-lpc32xx/phy3250.c
+ * Platform support for LPC32xx SoC
  *
  * Author: Kevin Wells <kevin.wells@nxp.com>
  *
+ * Copyright (C) 2012 Roland Stigge <stigge@antcom.de>
  * Copyright (C) 2010 NXP Semiconductors
  *
  * This program is free software; you can redistribute it and/or modify
@@ -25,11 +26,16 @@
 #include <linux/device.h>
 #include <linux/spi/spi.h>
 #include <linux/spi/eeprom.h>
-#include <linux/leds.h>
 #include <linux/gpio.h>
 #include <linux/amba/bus.h>
 #include <linux/amba/clcd.h>
 #include <linux/amba/pl022.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/of_irq.h>
+#include <linux/of_platform.h>
+#include <linux/clk.h>
+#include <linux/amba/pl08x.h>
 
 #include <asm/setup.h>
 #include <asm/mach-types.h>
@@ -47,7 +53,6 @@
 #define SPI0_CS_GPIO	LPC32XX_GPIO(LPC32XX_GPIO_P3_GRP, 5)
 #define LCD_POWER_GPIO	LPC32XX_GPIO(LPC32XX_GPO_P3_GRP, 0)
 #define BKL_POWER_GPIO	LPC32XX_GPIO(LPC32XX_GPO_P3_GRP, 4)
-#define LED_GPIO	LPC32XX_GPIO(LPC32XX_GPO_P3_GRP, 1)
 
 /*
  * AMBA LCD controller
@@ -150,9 +155,6 @@ static struct clcd_board lpc32xx_clcd_da
 	.remove		= lpc32xx_clcd_remove,
 };
 
-static AMBA_AHB_DEVICE(lpc32xx_clcd, "dev:clcd", 0,
-	LPC32XX_LCD_BASE, { IRQ_LPC32XX_LCD }, &lpc32xx_clcd_data);
-
 /*
  * AMBA SSP (SPI)
  */
@@ -180,8 +182,11 @@ static struct pl022_ssp_controller lpc32
 	.enable_dma		= 0,
 };
 
-static AMBA_APB_DEVICE(lpc32xx_ssp0, "dev:ssp0", 0,
-	LPC32XX_SSP0_BASE, { IRQ_LPC32XX_SSP0 }, &lpc32xx_ssp0_data);
+static struct pl022_ssp_controller lpc32xx_ssp1_data = {
+	.bus_id			= 1,
+	.num_chipselect		= 1,
+	.enable_dma		= 0,
+};
 
 /* AT25 driver registration */
 static int __init phy3250_spi_board_register(void)
@@ -221,73 +226,33 @@ static int __init phy3250_spi_board_regi
 }
 arch_initcall(phy3250_spi_board_register);
 
-static struct i2c_board_info __initdata phy3250_i2c_board_info[] = {
-	{
-		I2C_BOARD_INFO("pcf8563", 0x51),
-	},
-};
-
-static struct gpio_led phy_leds[] = {
-	{
-		.name			= "led0",
-		.gpio			= LED_GPIO,
-		.active_low		= 1,
-		.default_trigger	= "heartbeat",
-	},
-};
-
-static struct gpio_led_platform_data led_data = {
-	.leds = phy_leds,
-	.num_leds = ARRAY_SIZE(phy_leds),
-};
-
-static struct platform_device lpc32xx_gpio_led_device = {
-	.name			= "leds-gpio",
-	.id			= -1,
-	.dev.platform_data	= &led_data,
-};
+static struct pl08x_platform_data pl08x_pd = {
 
-static struct platform_device *phy3250_devs[] __initdata = {
-	&lpc32xx_rtc_device,
-	&lpc32xx_tsc_device,
-	&lpc32xx_i2c0_device,
-	&lpc32xx_i2c1_device,
-	&lpc32xx_i2c2_device,
-	&lpc32xx_watchdog_device,
-	&lpc32xx_gpio_led_device,
-	&lpc32xx_adc_device,
-	&lpc32xx_ohci_device,
-	&lpc32xx_net_device,
 };
 
-static struct amba_device *amba_devs[] __initdata = {
-	&lpc32xx_clcd_device,
-	&lpc32xx_ssp0_device,
+static const struct of_dev_auxdata lpc32xx_auxdata_lookup[] __initconst = {
+	OF_DEV_AUXDATA("nxp,lpc32xx-adc", 0x40048000, "lpc32xx-adc", NULL),
+	OF_DEV_AUXDATA("nxp,lpc32xx-tsc", 0x40048000, "ts-lpc32xx", NULL),
+	OF_DEV_AUXDATA("nxp,pnx-i2c", 0x400A0000, "pnx-i2c.0", NULL),
+	OF_DEV_AUXDATA("nxp,pnx-i2c", 0x400A8000, "pnx-i2c.1", NULL),
+	OF_DEV_AUXDATA("nxp,pnx-i2c", 0x31020300, "pnx-i2c.2", NULL),
+	OF_DEV_AUXDATA("nxp,lpc32xx-rtc", 0x40024000, "lpc32xx_rtc", NULL),
+	OF_DEV_AUXDATA("nxp,lpc-eth", 0x31060000, "lpc-eth.0", NULL),
+	OF_DEV_AUXDATA("nxp,pnx4008-wdt", 0x4003C000, "pnx4008-watchdog", NULL),
+	OF_DEV_AUXDATA("nxp,lpc32xx-gpio", 0x40028000, "lpc32xx-gpio", NULL),
+	OF_DEV_AUXDATA("nxp,ohci-nxp", 0x31020000, "usb-ohci", NULL),
+	OF_DEV_AUXDATA("nxp,lpc32xx-udc", 0x31020000, "lpc32xx_udc", NULL),
+	OF_DEV_AUXDATA("arm,pl022", 0x20084000, "dev:ssp0", &lpc32xx_ssp0_data),
+	OF_DEV_AUXDATA("arm,pl022", 0x2008C000, "dev:ssp1", &lpc32xx_ssp1_data),
+	OF_DEV_AUXDATA("arm,pl110", 0x31040000, "dev:clcd", &lpc32xx_clcd_data),
+	OF_DEV_AUXDATA("arm,pl080", 0x31000000, "pl08xdmac", &pl08x_pd),
+	OF_DEV_AUXDATA("arm,pl180", 0x20098000, "dev:mmc0", NULL),
+	{ }
 };
 
-/*
- * Board specific functions
- */
-static void __init phy3250_board_init(void)
+static void __init lpc3250_machine_init(void)
 {
 	u32 tmp;
-	int i;
-
-	lpc32xx_gpio_init();
-
-	/* Register GPIOs used on this board */
-	if (gpio_request(SPI0_CS_GPIO, "spi0 cs"))
-		printk(KERN_ERR "Error requesting gpio %u",
-			SPI0_CS_GPIO);
-	else if (gpio_direction_output(SPI0_CS_GPIO, 1))
-		printk(KERN_ERR "Error setting gpio %u to output",
-			SPI0_CS_GPIO);
-
-	/* Setup network interface for RMII mode */
-	tmp = __raw_readl(LPC32XX_CLKPWR_MACCLK_CTRL);
-	tmp &= ~LPC32XX_CLKPWR_MACCTRL_PINS_MSK;
-	tmp |= LPC32XX_CLKPWR_MACCTRL_USE_RMII_PINS;
-	__raw_writel(tmp, LPC32XX_CLKPWR_MACCLK_CTRL);
 
 	/* Setup SLC NAND controller muxing */
 	__raw_writel(LPC32XX_CLKPWR_NANDCLK_SEL_SLC,
@@ -300,6 +265,12 @@ static void __init phy3250_board_init(vo
 	tmp |= LPC32XX_CLKPWR_LCDCTRL_LCDTYPE_TFT16;
 	__raw_writel(tmp, LPC32XX_CLKPWR_LCDCLK_CTRL);
 
+	/* Set up USB power */
+	tmp = __raw_readl(LPC32XX_CLKPWR_USB_CTRL);
+	tmp |= LPC32XX_CLKPWR_USBCTRL_HCLK_EN |
+		LPC32XX_CLKPWR_USBCTRL_USBI2C_EN;
+	__raw_writel(tmp, LPC32XX_CLKPWR_USB_CTRL);
+
 	/* Set up I2C pull levels */
 	tmp = __raw_readl(LPC32XX_CLKPWR_I2C_CLK_CTRL);
 	tmp |= LPC32XX_CLKPWR_I2CCLK_USBI2CHI_DRIVE |
@@ -321,33 +292,35 @@ static void __init phy3250_board_init(vo
 	/*
 	 * AMBA peripheral clocks need to be enabled prior to AMBA device
 	 * detection or a data fault will occur, so enable the clocks
-	 * here. However, we don't want to enable them if the peripheral
-	 * isn't included in the image
+	 * here.
 	 */
-#ifdef CONFIG_FB_ARMCLCD
 	tmp = __raw_readl(LPC32XX_CLKPWR_LCDCLK_CTRL);
 	__raw_writel((tmp | LPC32XX_CLKPWR_LCDCTRL_CLK_EN),
 		LPC32XX_CLKPWR_LCDCLK_CTRL);
-#endif
-#ifdef CONFIG_SPI_PL022
+
 	tmp = __raw_readl(LPC32XX_CLKPWR_SSP_CLK_CTRL);
 	__raw_writel((tmp | LPC32XX_CLKPWR_SSPCTRL_SSPCLK0_EN),
 		LPC32XX_CLKPWR_SSP_CLK_CTRL);
-#endif
 
-	platform_add_devices(phy3250_devs, ARRAY_SIZE(phy3250_devs));
-	for (i = 0; i < ARRAY_SIZE(amba_devs); i++) {
-		struct amba_device *d = amba_devs[i];
-		amba_device_register(d, &iomem_resource);
-	}
+	tmp = __raw_readl(LPC32XX_CLKPWR_DMA_CLK_CTRL);
+	__raw_writel((tmp | LPC32XX_CLKPWR_DMACLKCTRL_CLK_EN),
+		     LPC32XX_CLKPWR_DMA_CLK_CTRL);
 
 	/* Test clock needed for UDA1380 initial init */
 	__raw_writel(LPC32XX_CLKPWR_TESTCLK2_SEL_MOSC |
 		LPC32XX_CLKPWR_TESTCLK_TESTCLK2_EN,
 		LPC32XX_CLKPWR_TEST_CLK_SEL);
 
-	i2c_register_board_info(0, phy3250_i2c_board_info,
-		ARRAY_SIZE(phy3250_i2c_board_info));
+	of_platform_populate(NULL, of_default_bus_match_table,
+			     lpc32xx_auxdata_lookup, NULL);
+
+	/* Register GPIOs used on this board */
+	if (gpio_request(SPI0_CS_GPIO, "spi0 cs"))
+		printk(KERN_ERR "Error requesting gpio %u",
+			SPI0_CS_GPIO);
+	else if (gpio_direction_output(SPI0_CS_GPIO, 1))
+		printk(KERN_ERR "Error setting gpio %u to output",
+			SPI0_CS_GPIO);
 }
 
 static int __init lpc32xx_display_uid(void)
@@ -363,12 +336,17 @@ static int __init lpc32xx_display_uid(vo
 }
 arch_initcall(lpc32xx_display_uid);
 
-MACHINE_START(PHY3250, "Phytec 3250 board with the LPC3250 Microcontroller")
-	/* Maintainer: Kevin Wells, NXP Semiconductors */
+static char const *lpc32xx_dt_compat[] __initdata = {
+	"nxp,lpc32xx",
+	NULL
+};
+
+DT_MACHINE_START(LPC32XX_DT, "LPC32XX SoC (Flattened Device Tree)")
 	.atag_offset	= 0x100,
 	.map_io		= lpc32xx_map_io,
 	.init_irq	= lpc32xx_init_irq,
 	.timer		= &lpc32xx_timer,
-	.init_machine	= phy3250_board_init,
+	.init_machine	= lpc3250_machine_init,
+	.dt_compat	= lpc32xx_dt_compat,
 	.restart	= lpc23xx_restart,
 MACHINE_END

Re: [PATCH v2 8/8] ARM: LPC32xx: Device tree support

[Thierry Reding]

[-- Attachment #1: Type: text/plain, Size: 4495 bytes --]

* Roland Stigge wrote:
> This patch does the actual device tree switch for the LPC32xx SoC.
> 
> Signed-off-by: Roland Stigge <stigge@antcom.de>
> 
> ---
> 
> Applies to v3.4-rc3
> 
>  Documentation/devicetree/bindings/arm/lpc32xx-mic.txt |   38 +++
>  Documentation/devicetree/bindings/arm/lpc32xx.txt     |    8 
>  arch/arm/Kconfig                                      |    1 
>  arch/arm/mach-lpc32xx/Kconfig                         |   26 --
>  arch/arm/mach-lpc32xx/clock.c                         |   77 +++----
>  arch/arm/mach-lpc32xx/common.c                        |  192 ------------------
>  arch/arm/mach-lpc32xx/common.h                        |   14 -
>  arch/arm/mach-lpc32xx/irq.c                           |   78 +++++--

Could this perhaps be split into another patch. Basically this is a
conversion to IRQ domain *and* device tree support. But maybe it isn't worth
the effort.

>  arch/arm/mach-lpc32xx/phy3250.c                       |  146 +++++--------

While at it, this should probably be renamed board-dt.c or something similar
since it is no longer phy3250 specific.

[...]
> --- linux-2.6.orig/arch/arm/mach-lpc32xx/phy3250.c
> +++ linux-2.6/arch/arm/mach-lpc32xx/phy3250.c
[...]
> -static void __init phy3250_board_init(void)
> +static void __init lpc3250_machine_init(void)
>  {
>  	u32 tmp;
> -	int i;
> -
> -	lpc32xx_gpio_init();
> -
> -	/* Register GPIOs used on this board */
> -	if (gpio_request(SPI0_CS_GPIO, "spi0 cs"))
> -		printk(KERN_ERR "Error requesting gpio %u",
> -			SPI0_CS_GPIO);
> -	else if (gpio_direction_output(SPI0_CS_GPIO, 1))
> -		printk(KERN_ERR "Error setting gpio %u to output",
> -			SPI0_CS_GPIO);
> -
> -	/* Setup network interface for RMII mode */
> -	tmp = __raw_readl(LPC32XX_CLKPWR_MACCLK_CTRL);
> -	tmp &= ~LPC32XX_CLKPWR_MACCTRL_PINS_MSK;
> -	tmp |= LPC32XX_CLKPWR_MACCTRL_USE_RMII_PINS;
> -	__raw_writel(tmp, LPC32XX_CLKPWR_MACCLK_CTRL);
>  
>  	/* Setup SLC NAND controller muxing */
>  	__raw_writel(LPC32XX_CLKPWR_NANDCLK_SEL_SLC,
> @@ -300,6 +265,12 @@ static void __init phy3250_board_init(vo
>  	tmp |= LPC32XX_CLKPWR_LCDCTRL_LCDTYPE_TFT16;
>  	__raw_writel(tmp, LPC32XX_CLKPWR_LCDCLK_CTRL);
>  
> +	/* Set up USB power */
> +	tmp = __raw_readl(LPC32XX_CLKPWR_USB_CTRL);
> +	tmp |= LPC32XX_CLKPWR_USBCTRL_HCLK_EN |
> +		LPC32XX_CLKPWR_USBCTRL_USBI2C_EN;
> +	__raw_writel(tmp, LPC32XX_CLKPWR_USB_CTRL);
> +
>  	/* Set up I2C pull levels */
>  	tmp = __raw_readl(LPC32XX_CLKPWR_I2C_CLK_CTRL);
>  	tmp |= LPC32XX_CLKPWR_I2CCLK_USBI2CHI_DRIVE |
> @@ -321,33 +292,35 @@ static void __init phy3250_board_init(vo
>  	/*
>  	 * AMBA peripheral clocks need to be enabled prior to AMBA device
>  	 * detection or a data fault will occur, so enable the clocks
> -	 * here. However, we don't want to enable them if the peripheral
> -	 * isn't included in the image
> +	 * here.
>  	 */
> -#ifdef CONFIG_FB_ARMCLCD
>  	tmp = __raw_readl(LPC32XX_CLKPWR_LCDCLK_CTRL);
>  	__raw_writel((tmp | LPC32XX_CLKPWR_LCDCTRL_CLK_EN),
>  		LPC32XX_CLKPWR_LCDCLK_CTRL);
> -#endif
> -#ifdef CONFIG_SPI_PL022
> +
>  	tmp = __raw_readl(LPC32XX_CLKPWR_SSP_CLK_CTRL);
>  	__raw_writel((tmp | LPC32XX_CLKPWR_SSPCTRL_SSPCLK0_EN),
>  		LPC32XX_CLKPWR_SSP_CLK_CTRL);
> -#endif
>  
> -	platform_add_devices(phy3250_devs, ARRAY_SIZE(phy3250_devs));
> -	for (i = 0; i < ARRAY_SIZE(amba_devs); i++) {
> -		struct amba_device *d = amba_devs[i];
> -		amba_device_register(d, &iomem_resource);
> -	}
> +	tmp = __raw_readl(LPC32XX_CLKPWR_DMA_CLK_CTRL);
> +	__raw_writel((tmp | LPC32XX_CLKPWR_DMACLKCTRL_CLK_EN),
> +		     LPC32XX_CLKPWR_DMA_CLK_CTRL);
>  	/* Test clock needed for UDA1380 initial init */
>  	__raw_writel(LPC32XX_CLKPWR_TESTCLK2_SEL_MOSC |
>  		LPC32XX_CLKPWR_TESTCLK_TESTCLK2_EN,
>  		LPC32XX_CLKPWR_TEST_CLK_SEL);

A lot of these seem to be gratuitous here. Can control of these clocks not be
exposed via the clock framework? That would allow the controllers to only
activate them when actually needed. This could also be done in follow up
patches, though.

[...]
> +	/* Register GPIOs used on this board */
> +	if (gpio_request(SPI0_CS_GPIO, "spi0 cs"))
> +		printk(KERN_ERR "Error requesting gpio %u",
> +			SPI0_CS_GPIO);
> +	else if (gpio_direction_output(SPI0_CS_GPIO, 1))
> +		printk(KERN_ERR "Error setting gpio %u to output",
> +			SPI0_CS_GPIO);

This should be initialized based on data from the device tree.

Thierry

[-- Attachment #2: Type: application/pgp-signature, Size: 198 bytes --]

Re: [PATCH v2 8/8] ARM: LPC32xx: Device tree support

[Arnd Bergmann]

On Wednesday 18 April 2012, Thierry Reding wrote:
> >  arch/arm/mach-lpc32xx/phy3250.c                       |  146 +++++--------
> 
> While at it, this should probably be renamed board-dt.c or something similar
> since it is no longer phy3250 specific.

I believe there are still a few phy3250 specific bits left in there, using auxdata,
right? In that case, I think it's fine to leave the file name for now, but I
think everyhting that is not phy3250 specific should just get moved into common.c
since we don't want to support any other board files for lpc32xx any more.

	Arnd

Re: [PATCH v2 8/8] ARM: LPC32xx: Device tree support

[Roland Stigge]

Hi Thierry,

thanks for your suggestions.

On 04/18/2012 08:02 AM, Thierry Reding wrote:
>> arch/arm/mach-lpc32xx/irq.c                           |   78
>> +++++--
> 
> Could this perhaps be split into another patch. Basically this is
> a conversion to IRQ domain *and* device tree support. But maybe it
> isn't worth the effort.

I was thinking the same. :-) Specifically, using
irq_domain_add_legacy() already builds upon available DT nodes. DT for
irq in turn seems to require irqdomain. So  would only split those two
on explicit request of reviewers when we have a reasonable solution to
this dilemma. Otherwise, I'd keep it in one patch.

>> arch/arm/mach-lpc32xx/phy3250.c                       |  146
>> +++++--------
> 
> While at it, this should probably be renamed board-dt.c or
> something similar since it is no longer phy3250 specific.

As Arnd wrote, there are board specific bits left (open questions
about how to DT'ize spi's controller_data, and DT-conversion of AMBA
CLCD). However, I will move board-independent stuff over to common.c.
(Keeping lpc32xx_auxdata_lookup[] in phy3250.c due to remaining
platform_data, though.)

>> -	platform_add_devices(phy3250_devs, ARRAY_SIZE(phy3250_devs)); -
>> for (i = 0; i < ARRAY_SIZE(amba_devs); i++) { -		struct
>> amba_device *d = amba_devs[i]; -		amba_device_register(d,
>> &iomem_resource); -	} +	tmp =
>> __raw_readl(LPC32XX_CLKPWR_DMA_CLK_CTRL); +	__raw_writel((tmp |
>> LPC32XX_CLKPWR_DMACLKCTRL_CLK_EN), +
>> LPC32XX_CLKPWR_DMA_CLK_CTRL); /* Test clock needed for UDA1380
>> initial init */ __raw_writel(LPC32XX_CLKPWR_TESTCLK2_SEL_MOSC | 
>> LPC32XX_CLKPWR_TESTCLK_TESTCLK2_EN, 
>> LPC32XX_CLKPWR_TEST_CLK_SEL);
> 
> A lot of these seem to be gratuitous here. Can control of these
> clocks not be exposed via the clock framework? That would allow the
> controllers to only activate them when actually needed. This could
> also be done in follow up patches, though.

Correct, clock framework is on my list of following patches.

> 
> [...]
>> +	/* Register GPIOs used on this board */ +	if
>> (gpio_request(SPI0_CS_GPIO, "spi0 cs")) +		printk(KERN_ERR "Error
>> requesting gpio %u", +			SPI0_CS_GPIO); +	else if
>> (gpio_direction_output(SPI0_CS_GPIO, 1)) +		printk(KERN_ERR
>> "Error setting gpio %u to output", +			SPI0_CS_GPIO);
> 
> This should be initialized based on data from the device tree.

Correct. This will be done in the later patch regarding spi controller
DT conversion. Still discussing this with Grant. For now, I'm
considering the above hunk as just moving around the resp. lines
without change. Please let me keep it until the SPI issue is solved.

Thanks,

Roland
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v2 8/8] ARM: LPC32xx: Device tree support

[Arnd Bergmann]

On Wednesday 18 April 2012, Roland Stigge wrote:
> As Arnd wrote, there are board specific bits left (open questions
> about how to DT'ize spi's controller_data, and DT-conversion of AMBA
> CLCD). However, I will move board-independent stuff over to common.c.
> (Keeping lpc32xx_auxdata_lookup[] in phy3250.c due to remaining
> platform_data, though.)

FWIW, I think you can remove the most of the auxdata table if you
change the clock lookup table to refer to the automatic device
names, e.g. 31020000.usbd instead of lpc32xx_udc.

	Arnd

Re: [PATCH v2 0/8] Second patchset for LPC32xx device tree conversion

[Dan Carpenter]

On Tue, Apr 17, 2012 at 07:08:19PM +0200, Roland Stigge wrote:
> Applies to v3.4-rc3
> 

This probably applies fine (the previous version did a couple days
ago), but it's always best to submit patches against linux-next.
The 3.4 kernel is in -rc already so this is 3.5 material.

regards,
dan carpenter

Re: [PATCH v2 0/8] Second patchset for LPC32xx device tree conversion

[Roland Stigge]

On 04/17/2012 10:50 PM, Dan Carpenter wrote:
> On Tue, Apr 17, 2012 at 07:08:19PM +0200, Roland Stigge wrote:
>> Applies to v3.4-rc3
>>
> 
> This probably applies fine (the previous version did a couple days
> ago), but it's always best to submit patches against linux-next.
> The 3.4 kernel is in -rc already so this is 3.5 material.

IIRC, I've been asked by Arnd to always build upon current rc tags.
(Sounded reasonable, since this one will become a stable release upon
which "my" patches will eventually be merged in the following merge window.)

Please point me to what I got wrong where, or try to agree on one
strategy here.

Thanks in advance,

Roland

Re: [PATCH v2 0/8] Second patchset for LPC32xx device tree conversion

[Arnd Bergmann]

On Tuesday 17 April 2012, Dan Carpenter wrote:
> On Tue, Apr 17, 2012 at 07:08:19PM +0200, Roland Stigge wrote:
> > Applies to v3.4-rc3
> > 
> 
> This probably applies fine (the previous version did a couple days
> ago), but it's always best to submit patches against linux-next.
> The 3.4 kernel is in -rc already so this is 3.5 material.

I disagree. The patches won't get applied on -next, they get applied
on an -rc release, so they should be submitted against that version
as well. I agree that it makes sense to test patches against -next
when there is reason to believe there might be conflicts, but it's
not mandatory. When you know about conflicts against other patches
that are already in -next, I suggest listing them in the cover 
letter (the patch 0/x) and suggest a resolution.

	Arnd

Re: [PATCH v2 0/8] Second patchset for LPC32xx device tree conversion

[Dan Carpenter]

On Wed, Apr 18, 2012 at 08:06:16AM +0000, Arnd Bergmann wrote:
> On Tuesday 17 April 2012, Dan Carpenter wrote:
> > On Tue, Apr 17, 2012 at 07:08:19PM +0200, Roland Stigge wrote:
> > > Applies to v3.4-rc3
> > > 
> > 
> > This probably applies fine (the previous version did a couple days
> > ago), but it's always best to submit patches against linux-next.
> > The 3.4 kernel is in -rc already so this is 3.5 material.
> 
> I disagree. The patches won't get applied on -next, they get applied
> on an -rc release, so they should be submitted against that version
> as well. I agree that it makes sense to test patches against -next
> when there is reason to believe there might be conflicts, but it's
> not mandatory. When you know about conflicts against other patches
> that are already in -next, I suggest listing them in the cover 
> letter (the patch 0/x) and suggest a resolution.
> 

I'm not sure I understand.  I thought everyone used the develop
against linux-next and backport the fixes model.  Are we going to
try merge these in 3.4?  It will still spend some time in linux-next
before we submit it, right?

To be honest, I mostly am familiar with staging/ where driver wide
white space cleanups are the norm.  Working against linux-next is
the only option for us or otherwise the conflicts would be too
much.

regards,
dan carpenter

Re: [PATCH v2 0/8] Second patchset for LPC32xx device tree conversion

[Arnd Bergmann]

On Wednesday 18 April 2012, Dan Carpenter wrote:
> I'm not sure I understand.  I thought everyone used the develop
> against linux-next and backport the fixes model.  Are we going to
> try merge these in 3.4?  It will still spend some time in linux-next
> before we submit it, right?

Correct, but the point is that you cannot add patches on top of -next
and have them included in a future -next version, because they never
merge cleanly.

The ideal workflow is:

1. develop on -rc
2. merge with latest -next, test and make sure it works there
3. submit for review against -rc
4. have patches included in -next once reviewed, but based on -rc
5. when merge window opens, have patches sent for upstream inclusion

The last two steps are slightly different for your own subsystem
compared to someone else's subsystem: In this case, Roland is the
author and sends the patches to me to have them included in -next
and I send them to Linus in the next merge window.

	Arnd

Re: [PATCH v2 0/8] Second patchset for LPC32xx device tree conversion

[Dan Carpenter]

Interesting.  Thanks for the explanation.

regards,
dan carpenter

Re: [PATCH v2 0/8] Second patchset for LPC32xx device tree conversion

[Mark Brown]

[-- Attachment #1: Type: text/plain, Size: 748 bytes --]

On Wed, Apr 18, 2012 at 11:00:39AM +0000, Arnd Bergmann wrote:

> 1. develop on -rc
> 2. merge with latest -next, test and make sure it works there
> 3. submit for review against -rc
> 4. have patches included in -next once reviewed, but based on -rc
> 5. when merge window opens, have patches sent for upstream inclusion

Steps 3 and 4 should be to submit against whatever branch is appropriate
for the subsystem and driver - if people follow this process they're
going to get bounced back by a fair proportion of maintainers, -rc isn't
universally what people are looking for so people should be aware that
they need to pay attention here.  

Generally I'd say the development version is a safer bet than -rc for
most subsystems.

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 836 bytes --]

Re: [PATCH v2 0/8] Second patchset for LPC32xx device tree conversion

[Arnd Bergmann]

On Wednesday 18 April 2012, Mark Brown wrote:
> On Wed, Apr 18, 2012 at 11:00:39AM +0000, Arnd Bergmann wrote:
> 
> > 1. develop on -rc
> > 2. merge with latest -next, test and make sure it works there
> > 3. submit for review against -rc
> > 4. have patches included in -next once reviewed, but based on -rc
> > 5. when merge window opens, have patches sent for upstream inclusion
> 
> Steps 3 and 4 should be to submit against whatever branch is appropriate
> for the subsystem and driver - if people follow this process they're
> going to get bounced back by a fair proportion of maintainers, -rc isn't
> universally what people are looking for so people should be aware that
> they need to pay attention here.  
> 
> Generally I'd say the development version is a safer bet than -rc for
> most subsystems.

Right. The description above was mostly done for the lpc32xx case,
which is going to get merged through the arm-soc tree and that doesn't
have a single development branch but instead has lots of them.

For subsystems that have just one branch, I agree that it makes sense
to develop against that one. Also for arm-soc, it can make sense
to base on one of the existing branches, but I prefer the default
to be to base on the -rc release so I can mix and match incoming
branches as needed.

	Arnd

Re: [PATCH v2 0/8] Second patchset for LPC32xx device tree conversion

[Mark Brown]

[-- Attachment #1: Type: text/plain, Size: 1607 bytes --]

On Wed, Apr 18, 2012 at 08:06:16AM +0000, Arnd Bergmann wrote:
> On Tuesday 17 April 2012, Dan Carpenter wrote:

> > This probably applies fine (the previous version did a couple days
> > ago), but it's always best to submit patches against linux-next.
> > The 3.4 kernel is in -rc already so this is 3.5 material.

> I disagree. The patches won't get applied on -next, they get applied
> on an -rc release, so they should be submitted against that version
> as well. I agree that it makes sense to test patches against -next
> when there is reason to believe there might be conflicts, but it's
> not mandatory. When you know about conflicts against other patches
> that are already in -next, I suggest listing them in the cover 
> letter (the patch 0/x) and suggest a resolution.

Dan's advice is generally sensible for most subsystems - it's not so
much that you should submit against -next itself but rather that you
should be checking what you're submitting against the latest versions of
the subsystems to make sure that you're up to date with the latest APIs
and best practices.  -next is (or should be) a good proxy for this.

If you know the relevant subsystem and driver are pretty stable then
it's generally not going to be an issue to just use the -rcs but there's
enough areas where there's rapid development of one form or another it's
likely to save you at least one round trip of "that's nice but needs
some updates because it won't apply/won't build/isn't using this new
feature we want to convert everyone to" that it's a good default to
check with -next first.

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 836 bytes --]