[PATCH 0/8] ARM: dts: OMAP5: ocp address space and few updates

[PATCH 0/8] ARM: dts: OMAP5: ocp address space and few updates

[Santosh Shilimkar]

Few updates for OMAP5 found during testing OMAP5 DT builds. Couple
of patches were already posted on the list. The series also contains
a patch which adds DMA request, IRQ lines and IO address space 
DT extraction support to hwmod code so that we can get rid of
such information from hwmod data files. It has been tested on OMAP5
and OMAP4 devices.

Rajendra Nayak (1):
  ARM: dts: OMAP5: Specify nonsecure PPI IRQ for arch timer

Santosh Shilimkar (7):
  ARM: dts: omap5-evm: Update available memory to 2032 MB
  ARM: dts: OMAP5: Align the local timer dt node as per the current
    binding code
  ARM: dts: OMAP5: Move the gic node out of ocp space
  ARM: dts: OMAP5: Update the timer and gic nodes for HYP kernel
    support
  ARM: dts: OMAP5: Update keypad reg property
  ARM: OMAP: hwmod: extract module address space from DT blob
  ARM: dts: OMAP5: Add ocp address space for all hwmods

 arch/arm/boot/dts/omap5-evm.dts  |    2 +-
 arch/arm/boot/dts/omap5.dtsi     |  282 +++++++++++++++++++++++++++++++++++---
 arch/arm/mach-omap2/omap_hwmod.c |   31 ++++-
 3 files changed, 293 insertions(+), 22 deletions(-)

-- 
1.7.9.5

[PATCH 1/8] ARM: dts: omap5-evm: Update available memory to 2032 MB

[Santosh Shilimkar]

On OMAP5 to detect invalid/bad memory accesses, 16MB of DDR is used as a trap.
Hence available memory for linux OS is 2032 MB on boards popullated with 2 GB
memory.

Cc: Benoit Cousson <b-cousson@ti.com>

Signed-off-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
---
 arch/arm/boot/dts/omap5-evm.dts |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/arm/boot/dts/omap5-evm.dts b/arch/arm/boot/dts/omap5-evm.dts
index 8722c15..b0e218a 100644
--- a/arch/arm/boot/dts/omap5-evm.dts
+++ b/arch/arm/boot/dts/omap5-evm.dts
@@ -16,7 +16,7 @@
 
 	memory {
 		device_type = "memory";
-		reg = <0x80000000 0x80000000>; /* 2 GB */
+		reg = <0x80000000 0x7F000000>; /* 2032 MB */
 	};
 
 	vmmcsd_fixed: fixedregulator-mmcsd {
-- 
1.7.9.5

[PATCH 2/8] ARM: dts: OMAP5: Align the local timer dt node as per the current binding code

[Santosh Shilimkar]

It has been decided to not duplicate banked modules dt nodes and that is
how the current arch timer dt extraction code is.

Update the OMAP5 dt file accordingly.

Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Benoit Cousson <b-cousson@ti.com>

Signed-off-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
---
 arch/arm/boot/dts/omap5.dtsi |   19 +++++++------------
 1 file changed, 7 insertions(+), 12 deletions(-)

diff --git a/arch/arm/boot/dts/omap5.dtsi b/arch/arm/boot/dts/omap5.dtsi
index 790bb2a..6609754 100644
--- a/arch/arm/boot/dts/omap5.dtsi
+++ b/arch/arm/boot/dts/omap5.dtsi
@@ -33,24 +33,19 @@
 	cpus {
 		cpu at 0 {
 			compatible = "arm,cortex-a15";
-			timer {
-				compatible = "arm,armv7-timer";
-				/* 14th PPI IRQ, active low level-sensitive */
-				interrupts = <1 14 0x308>;
-				clock-frequency = <6144000>;
-			};
 		};
 		cpu at 1 {
 			compatible = "arm,cortex-a15";
-			timer {
-				compatible = "arm,armv7-timer";
-				/* 14th PPI IRQ, active low level-sensitive */
-				interrupts = <1 14 0x308>;
-				clock-frequency = <6144000>;
-			};
 		};
 	};
 
+	timer {
+		compatible = "arm,armv7-timer";
+		/* 14th PPI IRQ, active low level-sensitive */
+		interrupts = <1 14 0x308>;
+		clock-frequency = <6144000>;
+	};
+
 	/*
 	 * The soc node represents the soc top level view. It is uses for IPs
 	 * that are not memory mapped in the MPU view or for the MPU itself.
-- 
1.7.9.5

[PATCH 3/8] ARM: dts: OMAP5: Specify nonsecure PPI IRQ for arch timer

[Santosh Shilimkar]

From: Rajendra Nayak <rnayak@ti.com>

Specify both secure as well as nonsecure PPI IRQ for arch
timer. This fixes the following errors seen on DT OMAP5 boot..

[    0.000000] arch_timer: No interrupt available, giving up

Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Benoit Cousson <b-cousson@ti.com>

Signed-off-by: Rajendra Nayak <rnayak@ti.com>
Signed-off-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
---
 arch/arm/boot/dts/omap5.dtsi |    5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/arch/arm/boot/dts/omap5.dtsi b/arch/arm/boot/dts/omap5.dtsi
index 6609754..4c9d984 100644
--- a/arch/arm/boot/dts/omap5.dtsi
+++ b/arch/arm/boot/dts/omap5.dtsi
@@ -41,8 +41,9 @@
 
 	timer {
 		compatible = "arm,armv7-timer";
-		/* 14th PPI IRQ, active low level-sensitive */
-		interrupts = <1 14 0x308>;
+		/* PPI secure/nonsecure IRQ, active low level-sensitive */
+		interrupts = <1 13 0x308>,
+			     <1 14 0x308>;
 		clock-frequency = <6144000>;
 	};
 
-- 
1.7.9.5

[PATCH 4/8] ARM: dts: OMAP5: Move the gic node out of ocp space

[Santosh Shilimkar]

GIC is not part of OCP space so move the gic dt node out of ocp
dt address space.

Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Benoit Cousson <b-cousson@ti.com>

Signed-off-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
---
 arch/arm/boot/dts/omap5.dtsi |   19 +++++++++++--------
 1 file changed, 11 insertions(+), 8 deletions(-)

diff --git a/arch/arm/boot/dts/omap5.dtsi b/arch/arm/boot/dts/omap5.dtsi
index 4c9d984..5b6b72c 100644
--- a/arch/arm/boot/dts/omap5.dtsi
+++ b/arch/arm/boot/dts/omap5.dtsi
@@ -18,6 +18,9 @@
 /include/ "skeleton.dtsi"
 
 / {
+	#address-cells = <1>;
+	#size-cells = <1>;
+
 	compatible = "ti,omap5";
 	interrupt-parent = <&gic>;
 
@@ -47,6 +50,14 @@
 		clock-frequency = <6144000>;
 	};
 
+	gic: interrupt-controller at 48211000 {
+		compatible = "arm,cortex-a15-gic";
+		interrupt-controller;
+		#interrupt-cells = <3>;
+		reg = <0x48211000 0x1000>,
+		      <0x48212000 0x1000>;
+	};
+
 	/*
 	 * The soc node represents the soc top level view. It is uses for IPs
 	 * that are not memory mapped in the MPU view or for the MPU itself.
@@ -96,14 +107,6 @@
 			pinctrl-single,function-mask = <0x7fff>;
 		};
 
-		gic: interrupt-controller at 48211000 {
-			compatible = "arm,cortex-a15-gic";
-			interrupt-controller;
-			#interrupt-cells = <3>;
-			reg = <0x48211000 0x1000>,
-			      <0x48212000 0x1000>;
-		};
-
 		gpio1: gpio at 4ae10000 {
 			compatible = "ti,omap4-gpio";
 			reg = <0x4ae10000 0x200>;
-- 
1.7.9.5

[PATCH 5/8] ARM: dts: OMAP5: Update the timer and gic nodes for HYP kernel support

[Santosh Shilimkar]

To be able to run kernel in HYP mode, virtual timer and gic node information
needs to be popullated.

Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Benoit Cousson <b-cousson@ti.com>

Signed-off-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
---
 arch/arm/boot/dts/omap5.dtsi |    8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/arch/arm/boot/dts/omap5.dtsi b/arch/arm/boot/dts/omap5.dtsi
index 5b6b72c..fa0c66b 100644
--- a/arch/arm/boot/dts/omap5.dtsi
+++ b/arch/arm/boot/dts/omap5.dtsi
@@ -46,7 +46,9 @@
 		compatible = "arm,armv7-timer";
 		/* PPI secure/nonsecure IRQ, active low level-sensitive */
 		interrupts = <1 13 0x308>,
-			     <1 14 0x308>;
+			     <1 14 0x308>,
+			     <1 11 0x308>,
+			     <1 10 0x308>;
 		clock-frequency = <6144000>;
 	};
 
@@ -55,7 +57,9 @@
 		interrupt-controller;
 		#interrupt-cells = <3>;
 		reg = <0x48211000 0x1000>,
-		      <0x48212000 0x1000>;
+		      <0x48212000 0x1000>,
+		      <0x48214000 0x2000>,
+		      <0x48216000 0x2000>;
 	};
 
 	/*
-- 
1.7.9.5

[PATCH 6/8] ARM: dts: OMAP5: Update keypad reg property

[Santosh Shilimkar]

Add missing OMAP keypad reg property information.

Cc: Benoit Cousson <b-cousson@ti.com>

Signed-off-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
---
 arch/arm/boot/dts/omap5.dtsi |    1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/arm/boot/dts/omap5.dtsi b/arch/arm/boot/dts/omap5.dtsi
index fa0c66b..269a328 100644
--- a/arch/arm/boot/dts/omap5.dtsi
+++ b/arch/arm/boot/dts/omap5.dtsi
@@ -335,6 +335,7 @@
 
 		keypad: keypad at 4ae1c000 {
 			compatible = "ti,omap4-keypad";
+			reg = <0x4ae1c000 0x400>;
 			ti,hwmods = "kbd";
 		};
 
-- 
1.7.9.5

[PATCH 7/8] ARM: OMAP: hwmod: extract module address space from DT blob

[Santosh Shilimkar]

Patch adds the code for extracting the module ocp address space from device
tree blob in case the hwmod address space look up fails.

The idea is to remove the address space data from hwmod and extract it from
DT blob.

Cc: Benoit Cousson <b-cousson@ti.com>

Signed-off-by: Rajendra Nayak <rnayak@ti.com>
Signed-off-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
---
 arch/arm/mach-omap2/omap_hwmod.c |   31 ++++++++++++++++++++++++++++++-
 1 file changed, 30 insertions(+), 1 deletion(-)

diff --git a/arch/arm/mach-omap2/omap_hwmod.c b/arch/arm/mach-omap2/omap_hwmod.c
index 4653efb..5aad348 100644
--- a/arch/arm/mach-omap2/omap_hwmod.c
+++ b/arch/arm/mach-omap2/omap_hwmod.c
@@ -138,6 +138,7 @@
 #include <linux/spinlock.h>
 #include <linux/slab.h>
 #include <linux/bootmem.h>
+#include <linux/of.h>
 
 #include "clock.h"
 #include "omap_hwmod.h"
@@ -2336,6 +2337,11 @@ static void __init _init_mpu_rt_base(struct omap_hwmod *oh, void *data)
 {
 	struct omap_hwmod_addr_space *mem;
 	void __iomem *va_start;
+	struct device_node *np;
+	const void *reg_prop;
+	const char *p;
+	unsigned long start = 0, size = 0;
+
 
 	if (!oh)
 		return;
@@ -2349,10 +2355,33 @@ static void __init _init_mpu_rt_base(struct omap_hwmod *oh, void *data)
 	if (!mem) {
 		pr_debug("omap_hwmod: %s: no MPU register target found\n",
 			 oh->name);
+
+		/* Extract the IO space from device tree blob */
+		if (!of_have_populated_dt())
+			return;
+
+		for_each_child_of_node(of_find_node_by_name(NULL, "ocp"), np) {
+			if (of_find_property(np, "ti,hwmods", NULL)) {
+				p = of_get_property(np, "ti,hwmods", NULL);
+				if (!strcmp(p, oh->name)) {
+					reg_prop = of_get_property(np, "reg",
+									NULL);
+					start = of_read_number(reg_prop, 1);
+					size = of_read_number(reg_prop + 4, 1);
+				}
+			}
+		}
+	} else {
+		start = mem->pa_start;
+		size = mem->pa_end - mem->pa_start;
+	}
+
+	if (!start) {
+		pr_debug("omap_hwmod: %s: No address space found\n", oh->name);
 		return;
 	}
 
-	va_start = ioremap(mem->pa_start, mem->pa_end - mem->pa_start);
+	va_start = ioremap(start, size);
 	if (!va_start) {
 		pr_err("omap_hwmod: %s: Could not ioremap\n", oh->name);
 		return;
-- 
1.7.9.5

[PATCH 8/8] ARM: dts: OMAP5: Add ocp address space for all hwmods

[Santosh Shilimkar]

Patch adds the OCP address space for all missing hwmod from existing
DT file. Note that the compatible isn't added by purpose to ensure that for
these hwmod, devices are not getting created.

Cc: Benoit Cousson <b-cousson@ti.com>

Signed-off-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
---
 arch/arm/boot/dts/omap5.dtsi |  238 ++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 238 insertions(+)

diff --git a/arch/arm/boot/dts/omap5.dtsi b/arch/arm/boot/dts/omap5.dtsi
index 269a328..915fd11 100644
--- a/arch/arm/boot/dts/omap5.dtsi
+++ b/arch/arm/boot/dts/omap5.dtsi
@@ -500,5 +500,243 @@
 			hw-caps-ll-interface;
 			hw-caps-temp-alert;
 		};
+
+		l4_emif_ocp_fw: l4_emif_ocp_fw at 4a20c000 {
+			reg = <0x4a20c000 0x100>;
+			ti,hwmods = "emif_ocp_fw";
+		};
+
+		mpu_l3_main_1: mpu_l3_main_1 at 44000000 {
+			reg = <0x44000000 0x2000>;
+			ti,hwmods = "l3_main_1";
+		};
+
+		mpu_l3_main_2: mpu_l3_main_2 at 44800000 {
+			reg = <0x44800000 0x3000>;
+			ti,hwmods = "l3_main_2";
+		};
+
+		mpu_l3_main_3: mpu_l3_main_3 at 45000000 {
+			reg = <0x44800000 0x4000>;
+			ti,hwmods = "l3_main_3";
+		};
+
+		l4_ocp_wp_noc: l4_ocp_wp_noc at 4a102000 {
+			reg = <0x44800000 0x80>;
+			ti,hwmods = "ocp_wp_noc";
+		};
+
+		abe_aess: abe_aess at 401f1000 {
+			reg = <0x44800000 0x400>;
+			ti,hwmods = "aess";
+		};
+
+		c2c: c2c at 5c000000 {
+			reg = <0x5c000000 0x100>;
+			ti,hwmods = "c2c";
+		};
+
+		ctrl_module_core: ctrl_module_core at 4a002000 {
+			reg = <0x4a002000 0x800>,
+			      <0x4a002800 0x1000>;
+			ti,hwmods = "ctrl_module_core";
+		};
+
+		ctrl_module_wkup: ctrl_module_wkup at 4a002000 {
+			reg = <0x4ae0c000 0x800>,
+			      <0x4ae0c800 0x800>;
+			ti,hwmods = "ctrl_module_wkup";
+		};
+
+		dss: dss at 58000000 {
+			reg = <0x58000000 0x80>;
+			ti,hwmods = "dss_core";
+		};
+
+		dss_dispc: dss_dispc at 58001000 {
+			reg = <0x58001000 0x1000>;
+			ti,hwmods = "dss_dispc";
+		};
+
+		dss_dsi1_a: dss_dsi1_a at 58004000 {
+			reg = <0x58004000 0x200>;
+			ti,hwmods = "dss_dsi1_a";
+		};
+
+		dss_dsi1_b: dss_dsi1_b at 58005000 {
+			reg = <0x58005000 0x200>;
+			ti,hwmods = "dss_dsi1_b";
+		};
+
+		dss_dsi1_c: dss_dsi1_c at 58009000 {
+			reg = <0x58009000 0x200>;
+			ti,hwmods = "dss_dsi1_c";
+		};
+
+		dss_hdmi: dss_hdmi at 58060000 {
+			reg = <0x58060000 0x19000>;
+			ti,hwmods = "dss_hdmi";
+		};
+
+		dss_rfbi: dss_rfbi at 580020000 {
+			reg = <0x58002000 0x100>;
+			ti,hwmods = "dss_rfbi";
+		};
+
+		elm: elm at 48078000 {
+			reg = <0x48078000 0x1000>;
+			ti,hwmods = "elm";
+		};
+
+		fdif: elm at 4a10a000 {
+			reg = <0x4a10a000 0x200>;
+			ti,hwmods = "fdif";
+		};
+
+		gpmc: gpmc at 50000000 {
+			reg = <0x50000000 0x400>;
+			ti,hwmods = "gpmc";
+		};
+
+		gpu: gpu at 5600fe00 {
+			reg = <0x5600fe00 0x2000>;
+			ti,hwmods = "gpu";
+		};
+
+
+		hdq1w: hdq1w at 480b2000 {
+			reg = <0x480b2000 0x20>;
+			ti,hwmods = "hdq1w";
+		};
+
+		hsi: hsi at 4a05a000 {
+			reg = <0x4a05a000 0x2000>;
+			ti,hwmods = "hsi";
+		};
+
+		ipu: ipu at 55082000 {
+			reg = <0x55082000 0x100>;
+			ti,hwmods = "ipu";
+		};
+
+		intc_ipu_c0: intc_ipu_c0 at 48211000 {
+			reg = <0x48211000 0x2000>;
+			ti,hwmods = "intc_ipu_c0";
+		};
+
+		intc_ipu_c1: intc_ipu_c1 at 48211000 {
+			reg = <0x48211000 0x2000>;
+			ti,hwmods = "intc_ipu_c1";
+		};
+
+		mmu_ipu: mmu_ipu at 55080800 {
+			reg = <0x55080800 0x1000>;
+			ti,hwmods = "mmu_ipu";
+		};
+
+		iss: iss at 52000000 {
+			reg = <0x52000000 0x20>;
+			ti,hwmods = "iss";
+		};
+
+		iva: iva at 5a05a400 {
+			reg = <0x5a05a400 0x80>;
+			ti,hwmods = "iva";
+		};
+
+		mailbox: mailbox at 4a0f4000 {
+			reg = <0x4a0f4000 0x200>;
+			ti,hwmods = "mailbox";
+		};
+
+		mcasp: mcasp at 40128000 {
+			reg = <0x40128000 0x400>;
+			ti,hwmods = "mcasp";
+		};
+
+		mcspi1: mcspi1 at 48098000 {
+			reg = <0x48098000 0x200>;
+			ti,hwmods = "mcspi1";
+		};
+
+		mcspi2: mcspi2 at 4809a000 {
+			reg = <0x4809a000 0x200>;
+			ti,hwmods = "mcspi2";
+		};
+
+		mcspi3: mcspi3 at 480b8000 {
+			reg = <0x480b8000 0x200>;
+			ti,hwmods = "mcspi3";
+		};
+
+		mcspi4: mcspi4 at 480ba000 {
+			reg = <0x480ba000 0x200>;
+			ti,hwmods = "mcspi4";
+		};
+
+		ocp2scp1: ocp2scp1 at 4a080000 {
+			reg = <0x4a080000 0x20>;
+			ti,hwmods = "ocp2scp1";
+		};
+
+		sata: sata at 4a141100 {
+			reg = <0x4a141100 0x08>;
+			ti,hwmods = "sata";
+		};
+
+		scrm: scrm at 4ae0a000 {
+			reg = <0x4ae0a000 0x800>;
+			ti,hwmods = "scrm";
+		};
+
+		slimbus1: slimbus1 at 4012c000 {
+			reg = <0x4012c000 0x400>;
+			ti,hwmods = "slimbus1";
+		};
+
+		smartreflex_core: smartreflex_core at 4a0dd000 {
+			reg = <0x4a0dd000 0x40>;
+			ti,hwmods = "smartreflex_core";
+		};
+
+		smartreflex_mm: smartreflex_mm at 4a0db000 {
+			reg = <0x4a0db000 0x40>;
+			ti,hwmods = "smartreflex_mm";
+		};
+
+		smartreflex_mpu: smartreflex_mpu at 4a0d9000 {
+			reg = <0x4a0d9000 0x40>;
+			ti,hwmods = "smartreflex_mpu";
+		};
+
+		spinlock: spinlock at 4a0f6000 {
+			reg = <0x4a0f6000 0x1000>;
+			ti,hwmods = "spinlock";
+		};
+
+		usb_host_hs: usb_host_hs at 4a064c00 {
+			reg = <0x4a064c00 0x100>;
+			ti,hwmods = "usb_host_hs";
+		};
+
+		usb_otg_ss: usb_otg_ss at 4a030000 {
+			reg = <0x4a030000 0x100>;
+			ti,hwmods = "usb_otg_ss";
+		};
+
+		usb_tll_hs: usb_tll_hs at 4a062800 {
+			reg = <0x4a062800 0x400>;
+			ti,hwmods = "usb_tll_hs";
+		};
+
+		wd_timer2: wd_timer2 at 4ae14000 {
+			reg = <0x4ae14000 0x80>;
+			ti,hwmods = "wd_timer2";
+		};
+
+		wd_timer3: wd_timer3 at 40130000 {
+			reg = <0x40130000 0x80>;
+			ti,hwmods = "wd_timer3";
+		};
 	};
 };
-- 
1.7.9.5

[PATCH 7/8] ARM: OMAP: hwmod: extract module address space from DT blob

[Felipe Balbi]

Hi,

On Wed, Feb 20, 2013 at 09:08:54PM +0530, Santosh Shilimkar wrote:
> Patch adds the code for extracting the module ocp address space from device
> tree blob in case the hwmod address space look up fails.

IMHO you should lookup on DT first and if that fails try hwmod ;-)

-- 
balbi
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20130220/716cae75/attachment.sig>

[PATCH 7/8] ARM: OMAP: hwmod: extract module address space from DT blob

[Santosh Shilimkar]

On Wednesday 20 February 2013 09:12 PM, Felipe Balbi wrote:
> Hi,
>
> On Wed, Feb 20, 2013 at 09:08:54PM +0530, Santosh Shilimkar wrote:
>> Patch adds the code for extracting the module ocp address space from device
>> tree blob in case the hwmod address space look up fails.
>
> IMHO you should lookup on DT first and if that fails try hwmod ;-)
>
I thought about that one. Since we have more devices rely on hwmod than
DT, I put the code that way. Eventually, we will need only the dt
version.

Regards,
Santosh

[PATCH 8/8] ARM: dts: OMAP5: Add ocp address space for all hwmods

[Benoit Cousson]

Hi Santosh,


On 02/20/2013 04:38 PM, Santosh Shilimkar wrote:
> Patch adds the OCP address space for all missing hwmod from existing
> DT file. Note that the compatible isn't added by purpose to ensure that for
> these hwmod, devices are not getting created.

Mmm, that will make the DTS a little bit weird and non-standard.
You'd better use the status=disabled flag for that.

> Cc: Benoit Cousson <b-cousson@ti.com>
> 
> Signed-off-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
> ---
>  arch/arm/boot/dts/omap5.dtsi |  238 ++++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 238 insertions(+)
> 
> diff --git a/arch/arm/boot/dts/omap5.dtsi b/arch/arm/boot/dts/omap5.dtsi
> index 269a328..915fd11 100644
> --- a/arch/arm/boot/dts/omap5.dtsi
> +++ b/arch/arm/boot/dts/omap5.dtsi
> @@ -500,5 +500,243 @@
>  			hw-caps-ll-interface;
>  			hw-caps-temp-alert;
>  		};
> +
> +		l4_emif_ocp_fw: l4_emif_ocp_fw at 4a20c000 {
> +			reg = <0x4a20c000 0x100>;
> +			ti,hwmods = "emif_ocp_fw";
> +		};

Are we doing anything useful with this entry, and potentially all the
other non-regular entries like that.

> +
> +		mpu_l3_main_1: mpu_l3_main_1 at 44000000 {
> +			reg = <0x44000000 0x2000>;
> +			ti,hwmods = "l3_main_1";
> +		};
> +
> +		mpu_l3_main_2: mpu_l3_main_2 at 44800000 {
> +			reg = <0x44800000 0x3000>;
> +			ti,hwmods = "l3_main_2";
> +		};
> +
> +		mpu_l3_main_3: mpu_l3_main_3 at 45000000 {
> +			reg = <0x44800000 0x4000>;
> +			ti,hwmods = "l3_main_3";
> +		};

The slit is due to PM partitioning, but that should be represented as an
unique device.

> +
> +		l4_ocp_wp_noc: l4_ocp_wp_noc at 4a102000 {
> +			reg = <0x44800000 0x80>;
> +			ti,hwmods = "ocp_wp_noc";
> +		};
> +
> +		abe_aess: abe_aess at 401f1000 {
> +			reg = <0x44800000 0x400>;
> +			ti,hwmods = "aess";
> +		};
> +
> +		c2c: c2c at 5c000000 {
> +			reg = <0x5c000000 0x100>;
> +			ti,hwmods = "c2c";
> +		};
> +
> +		ctrl_module_core: ctrl_module_core at 4a002000 {
> +			reg = <0x4a002000 0x800>,
> +			      <0x4a002800 0x1000>;
> +			ti,hwmods = "ctrl_module_core";
> +		};
> +
> +		ctrl_module_wkup: ctrl_module_wkup at 4a002000 {
> +			reg = <0x4ae0c000 0x800>,
> +			      <0x4ae0c800 0x800>;
> +			ti,hwmods = "ctrl_module_wkup";
> +		};

Should probably be one device as well.

> +
> +		dss: dss at 58000000 {
> +			reg = <0x58000000 0x80>;
> +			ti,hwmods = "dss_core";
> +		};
> +
> +		dss_dispc: dss_dispc at 58001000 {
> +			reg = <0x58001000 0x1000>;
> +			ti,hwmods = "dss_dispc";
> +		};
> +
> +		dss_dsi1_a: dss_dsi1_a at 58004000 {
> +			reg = <0x58004000 0x200>;
> +			ti,hwmods = "dss_dsi1_a";
> +		};
> +
> +		dss_dsi1_b: dss_dsi1_b at 58005000 {
> +			reg = <0x58005000 0x200>;
> +			ti,hwmods = "dss_dsi1_b";
> +		};
> +
> +		dss_dsi1_c: dss_dsi1_c at 58009000 {
> +			reg = <0x58009000 0x200>;
> +			ti,hwmods = "dss_dsi1_c";
> +		};
> +
> +		dss_hdmi: dss_hdmi at 58060000 {
> +			reg = <0x58060000 0x19000>;
> +			ti,hwmods = "dss_hdmi";
> +		};
> +
> +		dss_rfbi: dss_rfbi at 580020000 {
> +			reg = <0x58002000 0x100>;
> +			ti,hwmods = "dss_rfbi";
> +		};

For the DSS, you do have to respect the HW hierarchy. hwmod was flat,
and it was the source of a lot of PM dependency issue.

> +
> +		elm: elm at 48078000 {
> +			reg = <0x48078000 0x1000>;
> +			ti,hwmods = "elm";
> +		};
> +
> +		fdif: elm at 4a10a000 {
> +			reg = <0x4a10a000 0x200>;
> +			ti,hwmods = "fdif";
> +		};

Same issue than DSS. FDIF, ISS are somehow related.

> +
> +		gpmc: gpmc at 50000000 {
> +			reg = <0x50000000 0x400>;
> +			ti,hwmods = "gpmc";
> +		};
> +
> +		gpu: gpu at 5600fe00 {
> +			reg = <0x5600fe00 0x2000>;
> +			ti,hwmods = "gpu";
> +		};
> +
> +
> +		hdq1w: hdq1w at 480b2000 {
> +			reg = <0x480b2000 0x20>;
> +			ti,hwmods = "hdq1w";
> +		};
> +
> +		hsi: hsi at 4a05a000 {
> +			reg = <0x4a05a000 0x2000>;
> +			ti,hwmods = "hsi";
> +		};
> +
> +		ipu: ipu at 55082000 {
> +			reg = <0x55082000 0x100>;
> +			ti,hwmods = "ipu";
> +		};
> +
> +		intc_ipu_c0: intc_ipu_c0 at 48211000 {
> +			reg = <0x48211000 0x2000>;
> +			ti,hwmods = "intc_ipu_c0";
> +		};
> +
> +		intc_ipu_c1: intc_ipu_c1 at 48211000 {
> +			reg = <0x48211000 0x2000>;
> +			ti,hwmods = "intc_ipu_c1";
> +		};

Do we need such entries?

> +
> +		mmu_ipu: mmu_ipu at 55080800 {
> +			reg = <0x55080800 0x1000>;
> +			ti,hwmods = "mmu_ipu";
> +		};
> +
> +		iss: iss at 52000000 {
> +			reg = <0x52000000 0x20>;
> +			ti,hwmods = "iss";
> +		};
> +
> +		iva: iva at 5a05a400 {
> +			reg = <0x5a05a400 0x80>;
> +			ti,hwmods = "iva";
> +		};
> +
> +		mailbox: mailbox at 4a0f4000 {
> +			reg = <0x4a0f4000 0x200>;
> +			ti,hwmods = "mailbox";
> +		};
> +
> +		mcasp: mcasp at 40128000 {
> +			reg = <0x40128000 0x400>;
> +			ti,hwmods = "mcasp";
> +		};
> +
> +		mcspi1: mcspi1 at 48098000 {
> +			reg = <0x48098000 0x200>;
> +			ti,hwmods = "mcspi1";
> +		};
> +
> +		mcspi2: mcspi2 at 4809a000 {
> +			reg = <0x4809a000 0x200>;
> +			ti,hwmods = "mcspi2";
> +		};
> +
> +		mcspi3: mcspi3 at 480b8000 {
> +			reg = <0x480b8000 0x200>;
> +			ti,hwmods = "mcspi3";
> +		};
> +
> +		mcspi4: mcspi4 at 480ba000 {
> +			reg = <0x480ba000 0x200>;
> +			ti,hwmods = "mcspi4";
> +		};
> +
> +		ocp2scp1: ocp2scp1 at 4a080000 {
> +			reg = <0x4a080000 0x20>;
> +			ti,hwmods = "ocp2scp1";
> +		};
> +
> +		sata: sata at 4a141100 {
> +			reg = <0x4a141100 0x08>;
> +			ti,hwmods = "sata";
> +		};
> +
> +		scrm: scrm at 4ae0a000 {
> +			reg = <0x4ae0a000 0x800>;
> +			ti,hwmods = "scrm";
> +		};
> +
> +		slimbus1: slimbus1 at 4012c000 {
> +			reg = <0x4012c000 0x400>;
> +			ti,hwmods = "slimbus1";
> +		};
> +
> +		smartreflex_core: smartreflex_core at 4a0dd000 {
> +			reg = <0x4a0dd000 0x40>;
> +			ti,hwmods = "smartreflex_core";
> +		};
> +
> +		smartreflex_mm: smartreflex_mm at 4a0db000 {
> +			reg = <0x4a0db000 0x40>;
> +			ti,hwmods = "smartreflex_mm";
> +		};
> +
> +		smartreflex_mpu: smartreflex_mpu at 4a0d9000 {
> +			reg = <0x4a0d9000 0x40>;
> +			ti,hwmods = "smartreflex_mpu";
> +		};
> +
> +		spinlock: spinlock at 4a0f6000 {
> +			reg = <0x4a0f6000 0x1000>;
> +			ti,hwmods = "spinlock";
> +		};
> +
> +		usb_host_hs: usb_host_hs at 4a064c00 {
> +			reg = <0x4a064c00 0x100>;
> +			ti,hwmods = "usb_host_hs";
> +		};
> +
> +		usb_otg_ss: usb_otg_ss at 4a030000 {
> +			reg = <0x4a030000 0x100>;
> +			ti,hwmods = "usb_otg_ss";
> +		};
> +
> +		usb_tll_hs: usb_tll_hs at 4a062800 {
> +			reg = <0x4a062800 0x400>;
> +			ti,hwmods = "usb_tll_hs";
> +		};
> +
> +		wd_timer2: wd_timer2 at 4ae14000 {
> +			reg = <0x4ae14000 0x80>;
> +			ti,hwmods = "wd_timer2";
> +		};
> +
> +		wd_timer3: wd_timer3 at 40130000 {
> +			reg = <0x40130000 0x80>;
> +			ti,hwmods = "wd_timer3";
> +		};

In case of real device and existing binding, you must populate every
entries (IRQ, DMA...).

Regards,
Benoit

[PATCH 8/8] ARM: dts: OMAP5: Add ocp address space for all hwmods

[Santosh Shilimkar]

On Monday 25 February 2013 04:26 PM, Benoit Cousson wrote:
> Hi Santosh,
>
>
> On 02/20/2013 04:38 PM, Santosh Shilimkar wrote:
>> Patch adds the OCP address space for all missing hwmod from existing
>> DT file. Note that the compatible isn't added by purpose to ensure that for
>> these hwmod, devices are not getting created.
>
> Mmm, that will make the DTS a little bit weird and non-standard.
> You'd better use the status=disabled flag for that.
>
IIRC, we thought about it and then decided against it. I expect that
when a real device support get added the drivers will add compatible
information and also update the node with rest of the data.

>> Cc: Benoit Cousson <b-cousson@ti.com>
>>
>> Signed-off-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
>> ---
>>   arch/arm/boot/dts/omap5.dtsi |  238 ++++++++++++++++++++++++++++++++++++++++++
>>   1 file changed, 238 insertions(+)
>>
>> diff --git a/arch/arm/boot/dts/omap5.dtsi b/arch/arm/boot/dts/omap5.dtsi
>> index 269a328..915fd11 100644
>> --- a/arch/arm/boot/dts/omap5.dtsi
>> +++ b/arch/arm/boot/dts/omap5.dtsi
>> @@ -500,5 +500,243 @@
>>   			hw-caps-ll-interface;
>>   			hw-caps-temp-alert;
>>   		};
>> +
>> +		l4_emif_ocp_fw: l4_emif_ocp_fw at 4a20c000 {
>> +			reg = <0x4a20c000 0x100>;
>> +			ti,hwmods = "emif_ocp_fw";
>> +		};
>
> Are we doing anything useful with this entry, and potentially all the
> other non-regular entries like that.
>
Mostly not. Some of the entries did gave me trouble for the boot and
hence I retained as it is in hwmod. I will make another pass to see
if we can remove them.

>> +
>> +		mpu_l3_main_1: mpu_l3_main_1 at 44000000 {
>> +			reg = <0x44000000 0x2000>;
>> +			ti,hwmods = "l3_main_1";
>> +		};
>> +
>> +		mpu_l3_main_2: mpu_l3_main_2 at 44800000 {
>> +			reg = <0x44800000 0x3000>;
>> +			ti,hwmods = "l3_main_2";
>> +		};
>> +
>> +		mpu_l3_main_3: mpu_l3_main_3 at 45000000 {
>> +			reg = <0x44800000 0x4000>;
>> +			ti,hwmods = "l3_main_3";
>> +		};
>
> The slit is due to PM partitioning, but that should be represented as an
> unique device.
>
Can be done

>> +
>> +		l4_ocp_wp_noc: l4_ocp_wp_noc at 4a102000 {
>> +			reg = <0x44800000 0x80>;
>> +			ti,hwmods = "ocp_wp_noc";
>> +		};
>> +
>> +		abe_aess: abe_aess at 401f1000 {
>> +			reg = <0x44800000 0x400>;
>> +			ti,hwmods = "aess";
>> +		};
>> +
>> +		c2c: c2c at 5c000000 {
>> +			reg = <0x5c000000 0x100>;
>> +			ti,hwmods = "c2c";
>> +		};
>> +
>> +		ctrl_module_core: ctrl_module_core at 4a002000 {
>> +			reg = <0x4a002000 0x800>,
>> +			      <0x4a002800 0x1000>;
>> +			ti,hwmods = "ctrl_module_core";
>> +		};
>> +
>> +		ctrl_module_wkup: ctrl_module_wkup at 4a002000 {
>> +			reg = <0x4ae0c000 0x800>,
>> +			      <0x4ae0c800 0x800>;
>> +			ti,hwmods = "ctrl_module_wkup";
>> +		};
>
> Should probably be one device as well.
>
ok

>> +
>> +		dss: dss at 58000000 {
>> +			reg = <0x58000000 0x80>;
>> +			ti,hwmods = "dss_core";
>> +		};
>> +
>> +		dss_dispc: dss_dispc at 58001000 {
>> +			reg = <0x58001000 0x1000>;
>> +			ti,hwmods = "dss_dispc";
>> +		};
>> +
>> +		dss_dsi1_a: dss_dsi1_a at 58004000 {
>> +			reg = <0x58004000 0x200>;
>> +			ti,hwmods = "dss_dsi1_a";
>> +		};
>> +
>> +		dss_dsi1_b: dss_dsi1_b at 58005000 {
>> +			reg = <0x58005000 0x200>;
>> +			ti,hwmods = "dss_dsi1_b";
>> +		};
>> +
>> +		dss_dsi1_c: dss_dsi1_c at 58009000 {
>> +			reg = <0x58009000 0x200>;
>> +			ti,hwmods = "dss_dsi1_c";
>> +		};
>> +
>> +		dss_hdmi: dss_hdmi at 58060000 {
>> +			reg = <0x58060000 0x19000>;
>> +			ti,hwmods = "dss_hdmi";
>> +		};
>> +
>> +		dss_rfbi: dss_rfbi at 580020000 {
>> +			reg = <0x58002000 0x100>;
>> +			ti,hwmods = "dss_rfbi";
>> +		};
>
> For the DSS, you do have to respect the HW hierarchy. hwmod was flat,
> and it was the source of a lot of PM dependency issue.
>
I can update this but as I said, it is best to be left fro driver
owners since they know the IP and hierarchy best.


>> +
>> +		elm: elm at 48078000 {
>> +			reg = <0x48078000 0x1000>;
>> +			ti,hwmods = "elm";
>> +		};
>> +
>> +		fdif: elm at 4a10a000 {
>> +			reg = <0x4a10a000 0x200>;
>> +			ti,hwmods = "fdif";
>> +		};
>
> Same issue than DSS. FDIF, ISS are somehow related.
>
>> +
>> +		gpmc: gpmc at 50000000 {
>> +			reg = <0x50000000 0x400>;
>> +			ti,hwmods = "gpmc";
>> +		};
>> +
>> +		gpu: gpu at 5600fe00 {
>> +			reg = <0x5600fe00 0x2000>;
>> +			ti,hwmods = "gpu";
>> +		};
>> +
>> +
>> +		hdq1w: hdq1w at 480b2000 {
>> +			reg = <0x480b2000 0x20>;
>> +			ti,hwmods = "hdq1w";
>> +		};
>> +
>> +		hsi: hsi at 4a05a000 {
>> +			reg = <0x4a05a000 0x2000>;
>> +			ti,hwmods = "hsi";
>> +		};
>> +
>> +		ipu: ipu at 55082000 {
>> +			reg = <0x55082000 0x100>;
>> +			ti,hwmods = "ipu";
>> +		};
>> +
>> +		intc_ipu_c0: intc_ipu_c0 at 48211000 {
>> +			reg = <0x48211000 0x2000>;
>> +			ti,hwmods = "intc_ipu_c0";
>> +		};
>> +
>> +		intc_ipu_c1: intc_ipu_c1 at 48211000 {
>> +			reg = <0x48211000 0x2000>;
>> +			ti,hwmods = "intc_ipu_c1";
>> +		};
>
> Do we need such entries?
>
The point is if we remove it from here, we need to
remove associated hwmod entries as well o.w boot
will fail. Anyway, as mentioned I can do another
pass to see if I can remove some more.

For now what you see is what is in hwmod data file.

[..]

>> +		wd_timer2: wd_timer2 at 4ae14000 {
>> +			reg = <0x4ae14000 0x80>;
>> +			ti,hwmods = "wd_timer2";
>> +		};
>> +
>> +		wd_timer3: wd_timer3 at 40130000 {
>> +			reg = <0x40130000 0x80>;
>> +			ti,hwmods = "wd_timer3";
>> +		};
>
> In case of real device and existing binding, you must populate every
> entries (IRQ, DMA...).
>
I agree for the drivers which are already DT adapted. I
shall update those ones.

Thanks for review Benoit. I will update the patch based
on comments.

Regards,
Santosh

[PATCH 0/8] ARM: dts: OMAP5: ocp address space and few updates

[Santosh Shilimkar]

Benoit,

On Wednesday 20 February 2013 09:08 PM, Santosh Shilimkar wrote:
> Few updates for OMAP5 found during testing OMAP5 DT builds. Couple
> of patches were already posted on the list. The series also contains
> a patch which adds DMA request, IRQ lines and IO address space 
> DT extraction support to hwmod code so that we can get rid of
> such information from hwmod data files. It has been tested on OMAP5
> and OMAP4 devices.
> 
[..]

>   ARM: OMAP: hwmod: extract module address space from DT blob
Above patch even though dt extraction code touches the hwmod file.
Are you ok to keep this patch is the current series considering
both the patches needs to go through your tree. If it needs to
be in separate branch, then I can add that as part of the
other generic patches with your ack.

>  arch/arm/boot/dts/omap5-evm.dts  |    2 +-
>  arch/arm/boot/dts/omap5.dtsi     |  282 +++++++++++++++++++++++++++++++++++---
>  arch/arm/mach-omap2/omap_hwmod.c |   31 ++++-
>  3 files changed, 293 insertions(+), 22 deletions(-)
> 

Let me know your preference. Either way is fine by me.

Regards,
Santosh

[PATCH 0/8] ARM: dts: OMAP5: ocp address space and few updates

[Benoit Cousson]

Hi Santosh,

On 03/15/2013 11:24 AM, Santosh Shilimkar wrote:
> Benoit,
> 
> On Wednesday 20 February 2013 09:08 PM, Santosh Shilimkar wrote:
>> Few updates for OMAP5 found during testing OMAP5 DT builds. Couple
>> of patches were already posted on the list. The series also contains
>> a patch which adds DMA request, IRQ lines and IO address space 
>> DT extraction support to hwmod code so that we can get rid of
>> such information from hwmod data files. It has been tested on OMAP5
>> and OMAP4 devices.
>>
> [..]
> 
>>   ARM: OMAP: hwmod: extract module address space from DT blob
> Above patch even though dt extraction code touches the hwmod file.
> Are you ok to keep this patch is the current series considering
> both the patches needs to go through your tree. If it needs to
> be in separate branch, then I can add that as part of the
> other generic patches with your ack.
> 
>>  arch/arm/boot/dts/omap5-evm.dts  |    2 +-
>>  arch/arm/boot/dts/omap5.dtsi     |  282 +++++++++++++++++++++++++++++++++++---
>>  arch/arm/mach-omap2/omap_hwmod.c |   31 ++++-
>>  3 files changed, 293 insertions(+), 22 deletions(-)
>>
> 
> Let me know your preference. Either way is fine by me.

I can take both DTS and hwmod patches, so you can keep then in the same
series.

Thanks,
Benoit

[PATCH 0/8] ARM: dts: OMAP5: ocp address space and few updates

[Santosh Shilimkar]

On Friday 15 March 2013 06:15 PM, Benoit Cousson wrote:
> Hi Santosh,
> 
> On 03/15/2013 11:24 AM, Santosh Shilimkar wrote:
>> Benoit,
>>
>> On Wednesday 20 February 2013 09:08 PM, Santosh Shilimkar wrote:
>>> Few updates for OMAP5 found during testing OMAP5 DT builds. Couple
>>> of patches were already posted on the list. The series also contains
>>> a patch which adds DMA request, IRQ lines and IO address space 
>>> DT extraction support to hwmod code so that we can get rid of
>>> such information from hwmod data files. It has been tested on OMAP5
>>> and OMAP4 devices.
>>>
>> [..]
>>
>>>   ARM: OMAP: hwmod: extract module address space from DT blob
>> Above patch even though dt extraction code touches the hwmod file.
>> Are you ok to keep this patch is the current series considering
>> both the patches needs to go through your tree. If it needs to
>> be in separate branch, then I can add that as part of the
>> other generic patches with your ack.
>>
>>>  arch/arm/boot/dts/omap5-evm.dts  |    2 +-
>>>  arch/arm/boot/dts/omap5.dtsi     |  282 +++++++++++++++++++++++++++++++++++---
>>>  arch/arm/mach-omap2/omap_hwmod.c |   31 ++++-
>>>  3 files changed, 293 insertions(+), 22 deletions(-)
>>>
>>
>> Let me know your preference. Either way is fine by me.
> 
> I can take both DTS and hwmod patches, so you can keep then in the same
> series.
> 
Cool. I will wait for Jon to send you a branch and then post mine so
that I have one less dependency for testing.

Regards,
Santosh