Skip to contentSkip to conversation

OMAP: Add DT bindings to specify when devices should not be idled or reset

PATCH New Rajendra Nayak 5 participants 17 messages Oct 9, 2013 – Oct 15, 2013

Subsystems (6)
lore
Share & tools
lore
#0 lore raw

v2 cover OMAP: Add DT bindings to specify when devices should not be idled or reset

Changes in v2:
* Patches are split in such a way that DT and hwmod changes are kept in seperate patches so they can be pulled in by Benoit and Paul seperately.
* Binding names are slightly updated to address the concern from Tony and Benoit [1]
* Added a patch [5/5] to prevent GPIO reset on AM335x EVM-SK board, which was also the original motivation of this series since the board fails to boot from mainline despite support for AM335x and a am335x-evmsk.dts file existing in mainline for quite a while.

We have devices like co-processors, SoC internal busses, memory controllers etc which should not be idled or reset. In some cases erratas around IP blocks prevent them from either being idled or reset. Have a way to pass this information from Device tree, and get rid of similar information that exists as part of the omap_hwmod data files for various DT only OMAP SoCs.

Boot tested on OMAP4 panda es, AM335x EVM and AM335x EVM-SK

[1] http://www.mail-archive.com/linux-omap at vger.kernel.org/msg94349.html

Rajendra Nayak (5):
ARM: OMAP2+: hwmod: cleanup HWMOD_INIT_NO_RESET usage
ARM: dts: omap: Add new bindings for OMAP
ARM: OMAP2+: hwmod: Extract no-idle and no-reset info from DT
ARM: OMAP2+: hwmod: Cleanup usage of HWMOD_INIT_NO_RESET and
HWMOD_INIT_NO_IDLE
ARM: dts: AM335x-evmsk: Do not reset gpio0

 .../devicetree/bindings/arm/omap/omap.txt          |    3 ++-
 arch/arm/boot/dts/am335x-evmsk.dts                 |    4 ++++
 arch/arm/boot/dts/am33xx.dtsi                      |    2 ++
 arch/arm/boot/dts/omap4.dtsi                       |    3 +++
 arch/arm/boot/dts/omap5.dtsi                       |    2 ++
 arch/arm/mach-omap2/omap_hwmod.c                   |   23 ++++++++++++++------
 arch/arm/mach-omap2/omap_hwmod_33xx_data.c         |   22 +++++++++----------
 arch/arm/mach-omap2/omap_hwmod_44xx_data.c         |    6 ++---
 arch/arm/mach-omap2/omap_hwmod_54xx_data.c         |    4 +---
 9 files changed, 43 insertions(+), 26 deletions(-)
Signature · 1 line
1.7.9.5
#1 lore raw

PATCH v2 1/5 ARM: OMAP2+: hwmod: cleanup HWMOD_INIT_NO_RESET usage

For modules/IPs/hwmods which do not have -1- sys->class->reset()
and
-2- hardreset lines
and
-3- No way to do an ocp reset (no sysc control) the flag 'HWMOD_INIT_NO_RESET' is not much useful.

Cleanup all such instances across various hwmod data files.

---
 arch/arm/mach-omap2/omap_hwmod_33xx_data.c |   18 +++++++++---------
 arch/arm/mach-omap2/omap_hwmod_44xx_data.c |    6 +++---
 arch/arm/mach-omap2/omap_hwmod_54xx_data.c |    6 +++---
 3 files changed, 15 insertions(+), 15 deletions(-)
arch/arm/mach-omap2/omap_hwmod_33xx_data.c +9 -9
index 215894f..2815a91 100644
@@ -52,7 +52,7 @@ static struct omap_hwmod am33xx_emif_hwmod = {
 	.name		= "emif",
 	.class		= &am33xx_emif_hwmod_class,
 	.clkdm_name	= "l3_clkdm",
-	.flags		= (HWMOD_INIT_NO_IDLE | HWMOD_INIT_NO_RESET),
+	.flags		= HWMOD_INIT_NO_IDLE,
 	.main_clk	= "dpll_ddr_m2_div2_ck",
 	.prcm		= {
 		.omap4	= {
@@ -74,7 +74,7 @@ static struct omap_hwmod am33xx_l3_main_hwmod = {
 	.name		= "l3_main",
 	.class		= &am33xx_l3_hwmod_class,
 	.clkdm_name	= "l3_clkdm",
-	.flags		= (HWMOD_INIT_NO_IDLE | HWMOD_INIT_NO_RESET),
+	.flags		= HWMOD_INIT_NO_IDLE,
 	.main_clk	= "l3_gclk",
 	.prcm		= {
 		.omap4	= {
@@ -96,7 +96,7 @@ static struct omap_hwmod am33xx_l3_instr_hwmod = {
 	.name		= "l3_instr",
 	.class		= &am33xx_l3_hwmod_class,
 	.clkdm_name	= "l3_clkdm",
-	.flags		= (HWMOD_INIT_NO_IDLE | HWMOD_INIT_NO_RESET),
+	.flags		= HWMOD_INIT_NO_IDLE,
 	.main_clk	= "l3_gclk",
 	.prcm		= {
 		.omap4	= {
@@ -119,7 +119,7 @@ static struct omap_hwmod am33xx_l4_ls_hwmod = {
 	.name		= "l4_ls",
 	.class		= &am33xx_l4_hwmod_class,
 	.clkdm_name	= "l4ls_clkdm",
-	.flags		= (HWMOD_INIT_NO_IDLE | HWMOD_INIT_NO_RESET),
+	.flags		= HWMOD_INIT_NO_IDLE,
 	.main_clk	= "l4ls_gclk",
 	.prcm		= {
 		.omap4	= {
@@ -134,7 +134,7 @@ static struct omap_hwmod am33xx_l4_hs_hwmod = {
 	.name		= "l4_hs",
 	.class		= &am33xx_l4_hwmod_class,
 	.clkdm_name	= "l4hs_clkdm",
-	.flags		= (HWMOD_INIT_NO_IDLE | HWMOD_INIT_NO_RESET),
+	.flags		= HWMOD_INIT_NO_IDLE,
 	.main_clk	= "l4hs_gclk",
 	.prcm		= {
 		.omap4	= {
@@ -150,7 +150,7 @@ static struct omap_hwmod am33xx_l4_wkup_hwmod = {
 	.name		= "l4_wkup",
 	.class		= &am33xx_l4_hwmod_class,
 	.clkdm_name	= "l4_wkup_clkdm",
-	.flags		= (HWMOD_INIT_NO_IDLE | HWMOD_INIT_NO_RESET),
+	.flags		= HWMOD_INIT_NO_IDLE,
 	.prcm		= {
 		.omap4	= {
 			.clkctrl_offs	= AM33XX_CM_WKUP_L4WKUP_CLKCTRL_OFFSET,
@@ -170,7 +170,7 @@ static struct omap_hwmod am33xx_mpu_hwmod = {
 	.name		= "mpu",
 	.class		= &am33xx_mpu_hwmod_class,
 	.clkdm_name	= "mpu_clkdm",
-	.flags		= (HWMOD_INIT_NO_IDLE | HWMOD_INIT_NO_RESET),
+	.flags		= HWMOD_INIT_NO_IDLE,
 	.main_clk	= "dpll_mpu_m2_ck",
 	.prcm		= {
 		.omap4	= {
@@ -450,7 +450,7 @@ static struct omap_hwmod am33xx_ocmcram_hwmod = {
 	.name		= "ocmcram",
 	.class		= &am33xx_ocmcram_hwmod_class,
 	.clkdm_name	= "l3_clkdm",
-	.flags		= (HWMOD_INIT_NO_IDLE | HWMOD_INIT_NO_RESET),
+	.flags		= HWMOD_INIT_NO_IDLE,
 	.main_clk	= "l3_gclk",
 	.prcm		= {
 		.omap4	= {
@@ -532,7 +532,7 @@ static struct omap_hwmod am33xx_control_hwmod = {
 	.name		= "control",
 	.class		= &am33xx_control_hwmod_class,
 	.clkdm_name	= "l4_wkup_clkdm",
-	.flags		= (HWMOD_INIT_NO_IDLE | HWMOD_INIT_NO_RESET),
+	.flags		= HWMOD_INIT_NO_IDLE,
 	.main_clk	= "dpll_core_m4_div2_ck",
 	.prcm		= {
 		.omap4	= {
arch/arm/mach-omap2/omap_hwmod_44xx_data.c +3 -3
index 9c3b504..1e5b12c 100644
@@ -914,7 +914,7 @@ static struct omap_hwmod omap44xx_emif1_hwmod = {
 	.name		= "emif1",
 	.class		= &omap44xx_emif_hwmod_class,
 	.clkdm_name	= "l3_emif_clkdm",
-	.flags		= HWMOD_INIT_NO_IDLE | HWMOD_INIT_NO_RESET,
+	.flags		= HWMOD_INIT_NO_IDLE,
 	.main_clk	= "ddrphy_ck",
 	.prcm = {
 		.omap4 = {
@@ -930,7 +930,7 @@ static struct omap_hwmod omap44xx_emif2_hwmod = {
 	.name		= "emif2",
 	.class		= &omap44xx_emif_hwmod_class,
 	.clkdm_name	= "l3_emif_clkdm",
-	.flags		= HWMOD_INIT_NO_IDLE | HWMOD_INIT_NO_RESET,
+	.flags		= HWMOD_INIT_NO_IDLE,
 	.main_clk	= "ddrphy_ck",
 	.prcm = {
 		.omap4 = {
@@ -2193,7 +2193,7 @@ static struct omap_hwmod omap44xx_mpu_hwmod = {
 	.name		= "mpu",
 	.class		= &omap44xx_mpu_hwmod_class,
 	.clkdm_name	= "mpuss_clkdm",
-	.flags		= HWMOD_INIT_NO_IDLE | HWMOD_INIT_NO_RESET,
+	.flags		= HWMOD_INIT_NO_IDLE,
 	.main_clk	= "dpll_mpu_m2_ck",
 	.prcm = {
 		.omap4 = {
arch/arm/mach-omap2/omap_hwmod_54xx_data.c +3 -3
index cde4155..e3caee1 100644
@@ -352,7 +352,7 @@ static struct omap_hwmod omap54xx_emif1_hwmod = {
 	.name		= "emif1",
 	.class		= &omap54xx_emif_hwmod_class,
 	.clkdm_name	= "emif_clkdm",
-	.flags		= HWMOD_INIT_NO_IDLE | HWMOD_INIT_NO_RESET,
+	.flags		= HWMOD_INIT_NO_IDLE,
 	.main_clk	= "dpll_core_h11x2_ck",
 	.prcm = {
 		.omap4 = {
@@ -368,7 +368,7 @@ static struct omap_hwmod omap54xx_emif2_hwmod = {
 	.name		= "emif2",
 	.class		= &omap54xx_emif_hwmod_class,
 	.clkdm_name	= "emif_clkdm",
-	.flags		= HWMOD_INIT_NO_IDLE | HWMOD_INIT_NO_RESET,
+	.flags		= HWMOD_INIT_NO_IDLE,
 	.main_clk	= "dpll_core_h11x2_ck",
 	.prcm = {
 		.omap4 = {
@@ -1135,7 +1135,7 @@ static struct omap_hwmod omap54xx_mpu_hwmod = {
 	.name		= "mpu",
 	.class		= &omap54xx_mpu_hwmod_class,
 	.clkdm_name	= "mpu_clkdm",
-	.flags		= HWMOD_INIT_NO_IDLE | HWMOD_INIT_NO_RESET,
+	.flags		= HWMOD_INIT_NO_IDLE,
 	.main_clk	= "dpll_mpu_m2_ck",
 	.prcm = {
 		.omap4 = {
Signature · 1 line
1.7.9.5
maintainer
#2 lore raw

PATCH v2 1/5 ARM: OMAP2+: hwmod: cleanup HWMOD_INIT_NO_RESET usage

Quoted 11 lines · Rajendra Nayak

For modules/IPs/hwmods which do not have -1- sys->class->reset()
and
-2- hardreset lines
and
-3- No way to do an ocp reset (no sysc control) the flag 'HWMOD_INIT_NO_RESET' is not much useful.

Cleanup all such instances across various hwmod data files.

Thanks, queued.

- Paul

#3 lore raw

PATCH v2 2/5 ARM: dts: omap: Add new bindings for OMAP

In reply to Rajendra Nayak

On OMAP we have co-processor IPs, memory controllers, GPIOs which control regulators and power switches to PMIC, and SoC internal Bus IPs, some or most of which should either not be reset or idled or both at init. (In some cases there are erratas which prevent an IP from being reset)
Have a way to pass this information from DT.

Update the am33xx/omap4 and omap5 dtsi files with the new bindings for modules which either should not be idled. reset or both. A later patch would cleanup the same information that exists today as part of the hwmod data files.

---
 .../devicetree/bindings/arm/omap/omap.txt          |    3 ++-
 arch/arm/boot/dts/am33xx.dtsi                      |    2 ++
 arch/arm/boot/dts/omap4.dtsi                       |    3 +++
 arch/arm/boot/dts/omap5.dtsi                       |    2 ++
 4 files changed, 9 insertions(+), 1 deletion(-)
Documentation/devicetree/bindings/arm/omap/omap.txt +2 -1
index 91b7049..808c154 100644
@@ -21,7 +21,8 @@ Required properties:
 Optional properties:
 - ti,no_idle_on_suspend: When present, it prevents the PM to idle the module
   during suspend.
-
+- ti,no-reset-on-init: When present, the module should not be reset at init
+- ti,no-idle-on-init: When present, the module should not be idled at init
 
 Example:
 
arch/arm/boot/dts/am33xx.dtsi +2 -0
index f9c5da9..ec33ea0 100644
@@ -607,6 +607,7 @@
 			reg = <0x44d00000 0x4000	/* M3 UMEM */
 			       0x44d80000 0x2000>;	/* M3 DMEM */
 			ti,hwmods = "wkup_m3";
+			ti,no-reset-on-init;
 		};
 
 		elm: elm at 48080000 {
@@ -637,6 +638,7 @@
 		gpmc: gpmc at 50000000 {
 			compatible = "ti,am3352-gpmc";
 			ti,hwmods = "gpmc";
+			ti,no-idle-on-init;
 			reg = <0x50000000 0x2000>;
 			interrupts = <100>;
 			gpmc,num-cs = <7>;
arch/arm/boot/dts/omap4.dtsi +3 -0
index 22d9f2b..e8fe797 100644
@@ -214,6 +214,7 @@
 			gpmc,num-cs = <8>;
 			gpmc,num-waitpins = <4>;
 			ti,hwmods = "gpmc";
+			ti,no-idle-on-init;
 		};
 
 		uart1: serial at 4806a000 {
@@ -492,6 +493,7 @@
 			reg = <0x4c000000 0x100>;
 			interrupts = <GIC_SPI 110 IRQ_TYPE_LEVEL_HIGH>;
 			ti,hwmods = "emif1";
+			ti,no-idle-on-init;
 			phy-type = <1>;
 			hw-caps-read-idle-ctrl;
 			hw-caps-ll-interface;
@@ -503,6 +505,7 @@
 			reg = <0x4d000000 0x100>;
 			interrupts = <GIC_SPI 111 IRQ_TYPE_LEVEL_HIGH>;
 			ti,hwmods = "emif2";
+			ti,no-idle-on-init;
 			phy-type = <1>;
 			hw-caps-read-idle-ctrl;
 			hw-caps-ll-interface;
arch/arm/boot/dts/omap5.dtsi +2 -0
index 7cdea1b..a9d49df 100644
@@ -607,6 +607,7 @@
 		emif1: emif at 0x4c000000 {
 			compatible	= "ti,emif-4d5";
 			ti,hwmods	= "emif1";
+			ti,no-idle-on-init;
 			phy-type	= <2>; /* DDR PHY type: Intelli PHY */
 			reg = <0x4c000000 0x400>;
 			interrupts = <GIC_SPI 110 IRQ_TYPE_LEVEL_HIGH>;
@@ -618,6 +619,7 @@
 		emif2: emif at 0x4d000000 {
 			compatible	= "ti,emif-4d5";
 			ti,hwmods	= "emif2";
+			ti,no-idle-on-init;
 			phy-type	= <2>; /* DDR PHY type: Intelli PHY */
 			reg = <0x4d000000 0x400>;
 			interrupts = <GIC_SPI 111 IRQ_TYPE_LEVEL_HIGH>;
Signature · 1 line
1.7.9.5
#4 lore raw

PATCH v2 2/5 ARM: dts: omap: Add new bindings for OMAP

Hi Rajendra,

Quoted 7 lines · Rajendra Nayak

On OMAP we have co-processor IPs, memory controllers, GPIOs which control regulators and power switches to PMIC, and SoC internal Bus IPs, some or most of which should either not be reset or idled or both at init. (In some cases there are erratas which prevent an IP from being reset)
Have a way to pass this information from DT.

Did you get some acked-by from the DT maintainers? Every new bindings must be carefully reviewed now.

Regards,
Benoit

Quoted 98 lines

Update the am33xx/omap4 and omap5 dtsi files with the new bindings for modules which either should not be idled. reset or both. A later patch would cleanup the same information that exists today as part of the hwmod data files.

---
  .../devicetree/bindings/arm/omap/omap.txt          |    3 ++-
  arch/arm/boot/dts/am33xx.dtsi                      |    2 ++
  arch/arm/boot/dts/omap4.dtsi                       |    3 +++
  arch/arm/boot/dts/omap5.dtsi                       |    2 ++
  4 files changed, 9 insertions(+), 1 deletion(-)
Documentation/devicetree/bindings/arm/omap/omap.txt +2 -1
index 91b7049..808c154 100644
@@ -21,7 +21,8 @@ Required properties:
  Optional properties:
  - ti,no_idle_on_suspend: When present, it prevents the PM to idle the module
    during suspend.
-
+- ti,no-reset-on-init: When present, the module should not be reset at init
+- ti,no-idle-on-init: When present, the module should not be idled at init
 
  Example:
 
arch/arm/boot/dts/am33xx.dtsi +2 -0
index f9c5da9..ec33ea0 100644
@@ -607,6 +607,7 @@
  			reg = <0x44d00000 0x4000	/* M3 UMEM */
  			       0x44d80000 0x2000>;	/* M3 DMEM */
  			ti,hwmods = "wkup_m3";
+			ti,no-reset-on-init;
  		};
 
  		elm: elm at 48080000 {
@@ -637,6 +638,7 @@
  		gpmc: gpmc at 50000000 {
  			compatible = "ti,am3352-gpmc";
  			ti,hwmods = "gpmc";
+			ti,no-idle-on-init;
  			reg = <0x50000000 0x2000>;
  			interrupts = <100>;
  			gpmc,num-cs = <7>;
arch/arm/boot/dts/omap4.dtsi +3 -0
index 22d9f2b..e8fe797 100644
@@ -214,6 +214,7 @@
  			gpmc,num-cs = <8>;
  			gpmc,num-waitpins = <4>;
  			ti,hwmods = "gpmc";
+			ti,no-idle-on-init;
  		};
 
  		uart1: serial at 4806a000 {
@@ -492,6 +493,7 @@
  			reg = <0x4c000000 0x100>;
  			interrupts = <GIC_SPI 110 IRQ_TYPE_LEVEL_HIGH>;
  			ti,hwmods = "emif1";
+			ti,no-idle-on-init;
  			phy-type = <1>;
  			hw-caps-read-idle-ctrl;
  			hw-caps-ll-interface;
@@ -503,6 +505,7 @@
  			reg = <0x4d000000 0x100>;
  			interrupts = <GIC_SPI 111 IRQ_TYPE_LEVEL_HIGH>;
  			ti,hwmods = "emif2";
+			ti,no-idle-on-init;
  			phy-type = <1>;
  			hw-caps-read-idle-ctrl;
  			hw-caps-ll-interface;
arch/arm/boot/dts/omap5.dtsi +2 -0
index 7cdea1b..a9d49df 100644
@@ -607,6 +607,7 @@
  		emif1: emif at 0x4c000000 {
  			compatible	= "ti,emif-4d5";
  			ti,hwmods	= "emif1";
+			ti,no-idle-on-init;
  			phy-type	= <2>; /* DDR PHY type: Intelli PHY */
  			reg = <0x4c000000 0x400>;
  			interrupts = <GIC_SPI 110 IRQ_TYPE_LEVEL_HIGH>;
@@ -618,6 +619,7 @@
  		emif2: emif at 0x4d000000 {
  			compatible	= "ti,emif-4d5";
  			ti,hwmods	= "emif2";
+			ti,no-idle-on-init;
  			phy-type	= <2>; /* DDR PHY type: Intelli PHY */
  			reg = <0x4d000000 0x400>;
  			interrupts = <GIC_SPI 111 IRQ_TYPE_LEVEL_HIGH>;
#5 lore raw

PATCH v2 2/5 ARM: dts: omap: Add new bindings for OMAP

Quoted 11 lines · Benoit Cousson

Hi Rajendra,

Quoted 7 lines · Rajendra Nayak

On OMAP we have co-processor IPs, memory controllers, GPIOs which control regulators and power switches to PMIC, and SoC internal Bus IPs, some or most of which should either not be reset or idled or both at init. (In some cases there are erratas which prevent an IP from being reset)
Have a way to pass this information from DT.

Quoted 2 lines

Did you get some acked-by from the DT maintainers? Every new bindings must be carefully reviewed now.

Nope, I haven't. The patches are on the list for a couple of months now.

Quoted 103 lines

Regards,
Benoit

Quoted 98 lines

Update the am33xx/omap4 and omap5 dtsi files with the new bindings for modules which either should not be idled. reset or both. A later patch would cleanup the same information that exists today as part of the hwmod data files.

---
  .../devicetree/bindings/arm/omap/omap.txt          |    3 ++-
  arch/arm/boot/dts/am33xx.dtsi                      |    2 ++
  arch/arm/boot/dts/omap4.dtsi                       |    3 +++
  arch/arm/boot/dts/omap5.dtsi                       |    2 ++
  4 files changed, 9 insertions(+), 1 deletion(-)
Documentation/devicetree/bindings/arm/omap/omap.txt +2 -1
index 91b7049..808c154 100644
@@ -21,7 +21,8 @@ Required properties:
  Optional properties:
  - ti,no_idle_on_suspend: When present, it prevents the PM to idle the module
    during suspend.
-
+- ti,no-reset-on-init: When present, the module should not be reset at init
+- ti,no-idle-on-init: When present, the module should not be idled at init
 
  Example:
 
arch/arm/boot/dts/am33xx.dtsi +2 -0
index f9c5da9..ec33ea0 100644
@@ -607,6 +607,7 @@
              reg = <0x44d00000 0x4000    /* M3 UMEM */
                     0x44d80000 0x2000>;    /* M3 DMEM */
              ti,hwmods = "wkup_m3";
+            ti,no-reset-on-init;
          };
 
          elm: elm at 48080000 {
@@ -637,6 +638,7 @@
          gpmc: gpmc at 50000000 {
              compatible = "ti,am3352-gpmc";
              ti,hwmods = "gpmc";
+            ti,no-idle-on-init;
              reg = <0x50000000 0x2000>;
              interrupts = <100>;
              gpmc,num-cs = <7>;
arch/arm/boot/dts/omap4.dtsi +3 -0
index 22d9f2b..e8fe797 100644
@@ -214,6 +214,7 @@
              gpmc,num-cs = <8>;
              gpmc,num-waitpins = <4>;
              ti,hwmods = "gpmc";
+            ti,no-idle-on-init;
          };
 
          uart1: serial at 4806a000 {
@@ -492,6 +493,7 @@
              reg = <0x4c000000 0x100>;
              interrupts = <GIC_SPI 110 IRQ_TYPE_LEVEL_HIGH>;
              ti,hwmods = "emif1";
+            ti,no-idle-on-init;
              phy-type = <1>;
              hw-caps-read-idle-ctrl;
              hw-caps-ll-interface;
@@ -503,6 +505,7 @@
              reg = <0x4d000000 0x100>;
              interrupts = <GIC_SPI 111 IRQ_TYPE_LEVEL_HIGH>;
              ti,hwmods = "emif2";
+            ti,no-idle-on-init;
              phy-type = <1>;
              hw-caps-read-idle-ctrl;
              hw-caps-ll-interface;
arch/arm/boot/dts/omap5.dtsi +2 -0
index 7cdea1b..a9d49df 100644
@@ -607,6 +607,7 @@
          emif1: emif at 0x4c000000 {
              compatible    = "ti,emif-4d5";
              ti,hwmods    = "emif1";
+            ti,no-idle-on-init;
              phy-type    = <2>; /* DDR PHY type: Intelli PHY */
              reg = <0x4c000000 0x400>;
              interrupts = <GIC_SPI 110 IRQ_TYPE_LEVEL_HIGH>;
@@ -618,6 +619,7 @@
          emif2: emif at 0x4d000000 {
              compatible    = "ti,emif-4d5";
              ti,hwmods    = "emif2";
+            ti,no-idle-on-init;
              phy-type    = <2>; /* DDR PHY type: Intelli PHY */
              reg = <0x4d000000 0x400>;
              interrupts = <GIC_SPI 111 IRQ_TYPE_LEVEL_HIGH>;
#6 lore raw

PATCH v2 2/5 ARM: dts: omap: Add new bindings for OMAP

+ A couple of DT maintainers

Quoted 12 lines · Rajendra Nayak
Quoted 11 lines · Benoit Cousson

Hi Rajendra,

Quoted 7 lines · Rajendra Nayak

On OMAP we have co-processor IPs, memory controllers, GPIOs which control regulators and power switches to PMIC, and SoC internal Bus IPs, some or most of which should either not be reset or idled or both at init. (In some cases there are erratas which prevent an IP from being reset)
Have a way to pass this information from DT.

Quoted 5 lines
Quoted 2 lines

Did you get some acked-by from the DT maintainers? Every new bindings must be carefully reviewed now.

Nope, I haven't. The patches are on the list for a couple of months now.

Let's see if Mark and Rob are OK with that.

Thanks,
Benoit

Quoted 105 lines
Quoted 104 lines

Regards,
Benoit

Quoted 98 lines

Update the am33xx/omap4 and omap5 dtsi files with the new bindings for modules which either should not be idled. reset or both. A later patch would cleanup the same information that exists today as part of the hwmod data files.

---
   .../devicetree/bindings/arm/omap/omap.txt          |    3 ++-
   arch/arm/boot/dts/am33xx.dtsi                      |    2 ++
   arch/arm/boot/dts/omap4.dtsi                       |    3 +++
   arch/arm/boot/dts/omap5.dtsi                       |    2 ++
   4 files changed, 9 insertions(+), 1 deletion(-)
Documentation/devicetree/bindings/arm/omap/omap.txt +2 -1
index 91b7049..808c154 100644
@@ -21,7 +21,8 @@ Required properties:
   Optional properties:
   - ti,no_idle_on_suspend: When present, it prevents the PM to idle the module
     during suspend.
-
+- ti,no-reset-on-init: When present, the module should not be reset at init
+- ti,no-idle-on-init: When present, the module should not be idled at init
 
   Example:
 
arch/arm/boot/dts/am33xx.dtsi +2 -0
index f9c5da9..ec33ea0 100644
@@ -607,6 +607,7 @@
               reg = <0x44d00000 0x4000    /* M3 UMEM */
                      0x44d80000 0x2000>;    /* M3 DMEM */
               ti,hwmods = "wkup_m3";
+            ti,no-reset-on-init;
           };
 
           elm: elm at 48080000 {
@@ -637,6 +638,7 @@
           gpmc: gpmc at 50000000 {
               compatible = "ti,am3352-gpmc";
               ti,hwmods = "gpmc";
+            ti,no-idle-on-init;
               reg = <0x50000000 0x2000>;
               interrupts = <100>;
               gpmc,num-cs = <7>;
arch/arm/boot/dts/omap4.dtsi +3 -0
index 22d9f2b..e8fe797 100644
@@ -214,6 +214,7 @@
               gpmc,num-cs = <8>;
               gpmc,num-waitpins = <4>;
               ti,hwmods = "gpmc";
+            ti,no-idle-on-init;
           };
 
           uart1: serial at 4806a000 {
@@ -492,6 +493,7 @@
               reg = <0x4c000000 0x100>;
               interrupts = <GIC_SPI 110 IRQ_TYPE_LEVEL_HIGH>;
               ti,hwmods = "emif1";
+            ti,no-idle-on-init;
               phy-type = <1>;
               hw-caps-read-idle-ctrl;
               hw-caps-ll-interface;
@@ -503,6 +505,7 @@
               reg = <0x4d000000 0x100>;
               interrupts = <GIC_SPI 111 IRQ_TYPE_LEVEL_HIGH>;
               ti,hwmods = "emif2";
+            ti,no-idle-on-init;
               phy-type = <1>;
               hw-caps-read-idle-ctrl;
               hw-caps-ll-interface;
arch/arm/boot/dts/omap5.dtsi +2 -0
index 7cdea1b..a9d49df 100644
@@ -607,6 +607,7 @@
           emif1: emif at 0x4c000000 {
               compatible    = "ti,emif-4d5";
               ti,hwmods    = "emif1";
+            ti,no-idle-on-init;
               phy-type    = <2>; /* DDR PHY type: Intelli PHY */
               reg = <0x4c000000 0x400>;
               interrupts = <GIC_SPI 110 IRQ_TYPE_LEVEL_HIGH>;
@@ -618,6 +619,7 @@
           emif2: emif at 0x4d000000 {
               compatible    = "ti,emif-4d5";
               ti,hwmods    = "emif2";
+            ti,no-idle-on-init;
               phy-type    = <2>; /* DDR PHY type: Intelli PHY */
               reg = <0x4d000000 0x400>;
               interrupts = <GIC_SPI 111 IRQ_TYPE_LEVEL_HIGH>;
#7 lore raw

PATCH v2 2/5 ARM: dts: omap: Add new bindings for OMAP

Hi Rajendra,

Quoted 15 lines · Benoit Cousson

+ A couple of DT maintainers

Quoted 12 lines · Rajendra Nayak
Quoted 11 lines · Benoit Cousson

Hi Rajendra,

Quoted 7 lines · Rajendra Nayak

On OMAP we have co-processor IPs, memory controllers, GPIOs which control regulators and power switches to PMIC, and SoC internal Bus IPs, some or most of which should either not be reset or idled or both at init. (In some cases there are erratas which prevent an IP from being reset)
Have a way to pass this information from DT.

Quoted 7 lines
Quoted 5 lines
Quoted 2 lines

Did you get some acked-by from the DT maintainers? Every new bindings must be carefully reviewed now.

Nope, I haven't. The patches are on the list for a couple of months now.

Let's see if Mark and Rob are OK with that.

OK, I've just applied it after a small change in the subject:

ARM: dts: omap: Add reset/idle on init bindings for OMAP

Is that OK for you?

Thanks,
Benoit

Quoted 117 lines

Thanks,
Benoit

Quoted 112 lines
Quoted 111 lines

Regards,
Benoit

Quoted 105 lines

Update the am33xx/omap4 and omap5 dtsi files with the new bindings for modules which either should not be idled. reset or both. A later patch would cleanup the same information that exists today as part of the hwmod data files.

---
   .../devicetree/bindings/arm/omap/omap.txt          |    3 ++-
   arch/arm/boot/dts/am33xx.dtsi                      |    2 ++
   arch/arm/boot/dts/omap4.dtsi                       |    3 +++
   arch/arm/boot/dts/omap5.dtsi                       |    2 ++
   4 files changed, 9 insertions(+), 1 deletion(-)
Documentation/devicetree/bindings/arm/omap/omap.txt +0 -0
index 91b7049..808c154 100644
@@ -21,7 +21,8 @@ Required properties:
   Optional properties:
   - ti,no_idle_on_suspend: When present, it prevents the PM to idle

the module
during suspend.
-
+- ti,no-reset-on-init: When present, the module should not be reset at init
+- ti,no-idle-on-init: When present, the module should not be idled at init

Example:

arch/arm/boot/dts/am33xx.dtsi +2 -0
index f9c5da9..ec33ea0 100644
@@ -607,6 +607,7 @@
               reg = <0x44d00000 0x4000    /* M3 UMEM */
                      0x44d80000 0x2000>;    /* M3 DMEM */
               ti,hwmods = "wkup_m3";
+            ti,no-reset-on-init;
           };
 
           elm: elm at 48080000 {
@@ -637,6 +638,7 @@
           gpmc: gpmc at 50000000 {
               compatible = "ti,am3352-gpmc";
               ti,hwmods = "gpmc";
+            ti,no-idle-on-init;
               reg = <0x50000000 0x2000>;
               interrupts = <100>;
               gpmc,num-cs = <7>;
arch/arm/boot/dts/omap4.dtsi +3 -0
index 22d9f2b..e8fe797 100644
@@ -214,6 +214,7 @@
               gpmc,num-cs = <8>;
               gpmc,num-waitpins = <4>;
               ti,hwmods = "gpmc";
+            ti,no-idle-on-init;
           };
 
           uart1: serial at 4806a000 {
@@ -492,6 +493,7 @@
               reg = <0x4c000000 0x100>;
               interrupts = <GIC_SPI 110 IRQ_TYPE_LEVEL_HIGH>;
               ti,hwmods = "emif1";
+            ti,no-idle-on-init;
               phy-type = <1>;
               hw-caps-read-idle-ctrl;
               hw-caps-ll-interface;
@@ -503,6 +505,7 @@
               reg = <0x4d000000 0x100>;
               interrupts = <GIC_SPI 111 IRQ_TYPE_LEVEL_HIGH>;
               ti,hwmods = "emif2";
+            ti,no-idle-on-init;
               phy-type = <1>;
               hw-caps-read-idle-ctrl;
               hw-caps-ll-interface;
arch/arm/boot/dts/omap5.dtsi +2 -0
index 7cdea1b..a9d49df 100644
@@ -607,6 +607,7 @@
           emif1: emif at 0x4c000000 {
               compatible    = "ti,emif-4d5";
               ti,hwmods    = "emif1";
+            ti,no-idle-on-init;
               phy-type    = <2>; /* DDR PHY type: Intelli PHY */
               reg = <0x4c000000 0x400>;
               interrupts = <GIC_SPI 110 IRQ_TYPE_LEVEL_HIGH>;
@@ -618,6 +619,7 @@
           emif2: emif at 0x4d000000 {
               compatible    = "ti,emif-4d5";
               ti,hwmods    = "emif2";
+            ti,no-idle-on-init;
               phy-type    = <2>; /* DDR PHY type: Intelli PHY */
               reg = <0x4d000000 0x400>;
               interrupts = <GIC_SPI 111 IRQ_TYPE_LEVEL_HIGH>;
Signature · 4 lines
Beno?t Cousson
BayLibre
Embedded Linux Technology Lab
www.baylibre.com
#8 lore raw

PATCH v2 2/5 ARM: dts: omap: Add new bindings for OMAP

Quoted 24 lines · Benoit Cousson

Hi Rajendra,

Quoted 21 lines · Benoit Cousson

+ A couple of DT maintainers

Quoted 17 lines · Rajendra Nayak
Quoted 11 lines · Benoit Cousson

Hi Rajendra,

Quoted 7 lines · Rajendra Nayak

On OMAP we have co-processor IPs, memory controllers, GPIOs which control regulators and power switches to PMIC, and SoC internal Bus IPs, some or most of which should either not be reset or idled or both at init. (In some cases there are erratas which prevent an IP from being reset)
Have a way to pass this information from DT.

Quoted 2 lines

Did you get some acked-by from the DT maintainers? Every new bindings must be carefully reviewed now.

Nope, I haven't. The patches are on the list for a couple of months now.

Quoted 7 lines
Quoted 1 line

Let's see if Mark and Rob are OK with that.

OK, I've just applied it after a small change in the subject:

ARM: dts: omap: Add reset/idle on init bindings for OMAP

Is that OK for you?

Yes, sounds good. Thanks Benoit.

Quoted 123 lines

Thanks,
Benoit

Quoted 117 lines

Thanks,
Benoit

Quoted 112 lines
Quoted 111 lines

Regards,
Benoit

Quoted 105 lines

Update the am33xx/omap4 and omap5 dtsi files with the new bindings for modules which either should not be idled. reset or both. A later patch would cleanup the same information that exists today as part of the hwmod data files.

---
   .../devicetree/bindings/arm/omap/omap.txt          |    3 ++-
   arch/arm/boot/dts/am33xx.dtsi                      |    2 ++
   arch/arm/boot/dts/omap4.dtsi                       |    3 +++
   arch/arm/boot/dts/omap5.dtsi                       |    2 ++
   4 files changed, 9 insertions(+), 1 deletion(-)
Documentation/devicetree/bindings/arm/omap/omap.txt +0 -0
index 91b7049..808c154 100644
@@ -21,7 +21,8 @@ Required properties:
   Optional properties:
   - ti,no_idle_on_suspend: When present, it prevents the PM to idle

the module
during suspend.
-
+- ti,no-reset-on-init: When present, the module should not be reset at init
+- ti,no-idle-on-init: When present, the module should not be idled at init

Example:

arch/arm/boot/dts/am33xx.dtsi +2 -0
index f9c5da9..ec33ea0 100644
@@ -607,6 +607,7 @@
               reg = <0x44d00000 0x4000    /* M3 UMEM */
                      0x44d80000 0x2000>;    /* M3 DMEM */
               ti,hwmods = "wkup_m3";
+            ti,no-reset-on-init;
           };
 
           elm: elm at 48080000 {
@@ -637,6 +638,7 @@
           gpmc: gpmc at 50000000 {
               compatible = "ti,am3352-gpmc";
               ti,hwmods = "gpmc";
+            ti,no-idle-on-init;
               reg = <0x50000000 0x2000>;
               interrupts = <100>;
               gpmc,num-cs = <7>;
arch/arm/boot/dts/omap4.dtsi +3 -0
index 22d9f2b..e8fe797 100644
@@ -214,6 +214,7 @@
               gpmc,num-cs = <8>;
               gpmc,num-waitpins = <4>;
               ti,hwmods = "gpmc";
+            ti,no-idle-on-init;
           };
 
           uart1: serial at 4806a000 {
@@ -492,6 +493,7 @@
               reg = <0x4c000000 0x100>;
               interrupts = <GIC_SPI 110 IRQ_TYPE_LEVEL_HIGH>;
               ti,hwmods = "emif1";
+            ti,no-idle-on-init;
               phy-type = <1>;
               hw-caps-read-idle-ctrl;
               hw-caps-ll-interface;
@@ -503,6 +505,7 @@
               reg = <0x4d000000 0x100>;
               interrupts = <GIC_SPI 111 IRQ_TYPE_LEVEL_HIGH>;
               ti,hwmods = "emif2";
+            ti,no-idle-on-init;
               phy-type = <1>;
               hw-caps-read-idle-ctrl;
               hw-caps-ll-interface;
arch/arm/boot/dts/omap5.dtsi +2 -0
index 7cdea1b..a9d49df 100644
@@ -607,6 +607,7 @@
           emif1: emif at 0x4c000000 {
               compatible    = "ti,emif-4d5";
               ti,hwmods    = "emif1";
+            ti,no-idle-on-init;
               phy-type    = <2>; /* DDR PHY type: Intelli PHY */
               reg = <0x4c000000 0x400>;
               interrupts = <GIC_SPI 110 IRQ_TYPE_LEVEL_HIGH>;
@@ -618,6 +619,7 @@
           emif2: emif at 0x4d000000 {
               compatible    = "ti,emif-4d5";
               ti,hwmods    = "emif2";
+            ti,no-idle-on-init;
               phy-type    = <2>; /* DDR PHY type: Intelli PHY */
               reg = <0x4d000000 0x400>;
               interrupts = <GIC_SPI 111 IRQ_TYPE_LEVEL_HIGH>;
#9 lore raw

PATCH v2 3/5 ARM: OMAP2+: hwmod: Extract no-idle and no-reset info from DT

In reply to Rajendra Nayak

Now that we have DT bindings to specify which devices should not be reset and idled during init, make hwmod extract the information (and store them in internal flags) from Device tree.

---
 arch/arm/mach-omap2/omap_hwmod.c |   23 ++++++++++++++++-------
 1 file changed, 16 insertions(+), 7 deletions(-)
arch/arm/mach-omap2/omap_hwmod.c +16 -7
index d9ee0ff..b55923a 100644
@@ -2363,11 +2363,11 @@ static struct device_node *of_dev_hwmod_lookup(struct device_node *np,
  * are part of the device's address space can be ioremapped properly.
  * No return value.
  */
-static void __init _init_mpu_rt_base(struct omap_hwmod *oh, void *data)
+static void __init _init_mpu_rt_base(struct omap_hwmod *oh, void *data,
+				     struct device_node *np)
 {
 	struct omap_hwmod_addr_space *mem;
 	void __iomem *va_start = NULL;
-	struct device_node *np;
 
 	if (!oh)
 		return;
@@ -2383,12 +2383,10 @@ static void __init _init_mpu_rt_base(struct omap_hwmod *oh, void *data)
 			 oh->name);
 
 		/* Extract the IO space from device tree blob */
-		if (!of_have_populated_dt())
+		if (!np)
 			return;
 
-		np = of_dev_hwmod_lookup(of_find_node_by_name(NULL, "ocp"), oh);
-		if (np)
-			va_start = of_iomap(np, oh->mpu_rt_idx);
+		va_start = of_iomap(np, oh->mpu_rt_idx);
 	} else {
 		va_start = ioremap(mem->pa_start, mem->pa_end - mem->pa_start);
 	}
@@ -2420,12 +2418,16 @@ static void __init _init_mpu_rt_base(struct omap_hwmod *oh, void *data)
 static int __init _init(struct omap_hwmod *oh, void *data)
 {
 	int r;
+	struct device_node *np = NULL;
 
 	if (oh->_state != _HWMOD_STATE_REGISTERED)
 		return 0;
 
+	if (of_have_populated_dt())
+		np = of_dev_hwmod_lookup(of_find_node_by_name(NULL, "ocp"), oh);
+
 	if (oh->class->sysc)
-		_init_mpu_rt_base(oh, NULL);
+		_init_mpu_rt_base(oh, NULL, np);
 
 	r = _init_clocks(oh, NULL);
 	if (r < 0) {
@@ -2433,6 +2435,13 @@ static int __init _init(struct omap_hwmod *oh, void *data)
 		return -EINVAL;
 	}
 
+	if (np) {
+		if (of_find_property(np, "ti,no-reset-on-init", NULL))
+			oh->flags |= HWMOD_INIT_NO_RESET;
+		if (of_find_property(np, "ti,no-idle-on-init", NULL))
+			oh->flags |= HWMOD_INIT_NO_IDLE;
+	}
+
 	oh->_state = _HWMOD_STATE_INITIALIZED;
 
 	return 0;
Signature · 1 line
1.7.9.5
maintainer
#10 lore raw

PATCH v2 3/5 ARM: OMAP2+: hwmod: Extract no-idle and no-reset info from DT

Quoted 5 lines · Rajendra Nayak

Now that we have DT bindings to specify which devices should not be reset and idled during init, make hwmod extract the information (and store them in internal flags) from Device tree.

Thanks, queued.

- Paul

#11 lore raw

PATCH v2 4/5 ARM: OMAP2+: hwmod: Cleanup usage of HWMOD_INIT_NO_RESET and HWMOD_INIT_NO_IDLE

In reply to Rajendra Nayak

With DT bindings to specify which devices should not be idled and reset at init being in place, and the corresponding dtsi files for am33xx/omap4 and omap5 updated using those bindings, we can now clean up hwmod internal flags for HWMOD_INIT_NO_RESET and HWMOD_INIT_NO_IDLE which were infact used to specify the exact same information.

For GPMC, the HWMOD_INIT_NO_RESET flag seems to be added in hwmod not due to any errata around the GPMC IP, but rather because any timings set by the bootloader are not being correctly programmed by the kernel. This seems like something that needs to be fixed as part of GPMC driver in the kernel, and hence the flag is left as is in hwmod, which can be removed once the driver does what its expected to.

---
 arch/arm/mach-omap2/omap_hwmod_33xx_data.c |    4 ++--
 arch/arm/mach-omap2/omap_hwmod_44xx_data.c |    4 +---
 arch/arm/mach-omap2/omap_hwmod_54xx_data.c |    2 --
 3 files changed, 3 insertions(+), 7 deletions(-)
arch/arm/mach-omap2/omap_hwmod_33xx_data.c +2 -2
index 2815a91..e8dc72d 100644
@@ -198,7 +198,7 @@ static struct omap_hwmod am33xx_wkup_m3_hwmod = {
 	.class		= &am33xx_wkup_m3_hwmod_class,
 	.clkdm_name	= "l4_wkup_aon_clkdm",
 	/* Keep hardreset asserted */
-	.flags		= HWMOD_INIT_NO_RESET | HWMOD_NO_IDLEST,
+	.flags		= HWMOD_NO_IDLEST,
 	.main_clk	= "dpll_core_m4_div2_ck",
 	.prcm		= {
 		.omap4	= {
@@ -932,7 +932,7 @@ static struct omap_hwmod am33xx_gpmc_hwmod = {
 	.name		= "gpmc",
 	.class		= &am33xx_gpmc_hwmod_class,
 	.clkdm_name	= "l3s_clkdm",
-	.flags		= (HWMOD_INIT_NO_IDLE | HWMOD_INIT_NO_RESET),
+	.flags		= HWMOD_INIT_NO_RESET,
 	.main_clk	= "l3s_gclk",
 	.prcm		= {
 		.omap4	= {
arch/arm/mach-omap2/omap_hwmod_44xx_data.c +1 -3
index 1e5b12c..a507a70 100644
@@ -914,7 +914,6 @@ static struct omap_hwmod omap44xx_emif1_hwmod = {
 	.name		= "emif1",
 	.class		= &omap44xx_emif_hwmod_class,
 	.clkdm_name	= "l3_emif_clkdm",
-	.flags		= HWMOD_INIT_NO_IDLE,
 	.main_clk	= "ddrphy_ck",
 	.prcm = {
 		.omap4 = {
@@ -930,7 +929,6 @@ static struct omap_hwmod omap44xx_emif2_hwmod = {
 	.name		= "emif2",
 	.class		= &omap44xx_emif_hwmod_class,
 	.clkdm_name	= "l3_emif_clkdm",
-	.flags		= HWMOD_INIT_NO_IDLE,
 	.main_clk	= "ddrphy_ck",
 	.prcm = {
 		.omap4 = {
@@ -1184,7 +1182,7 @@ static struct omap_hwmod omap44xx_gpmc_hwmod = {
 	 * the kernel from the board file or DT data.
 	 * HWMOD_INIT_NO_RESET should be removed ASAP.
 	 */
-	.flags		= HWMOD_INIT_NO_IDLE | HWMOD_INIT_NO_RESET,
+	.flags		= HWMOD_INIT_NO_RESET,
 	.prcm = {
 		.omap4 = {
 			.clkctrl_offs = OMAP4_CM_L3_2_GPMC_CLKCTRL_OFFSET,
arch/arm/mach-omap2/omap_hwmod_54xx_data.c +0 -2
index e3caee1..e47f24d 100644
@@ -352,7 +352,6 @@ static struct omap_hwmod omap54xx_emif1_hwmod = {
 	.name		= "emif1",
 	.class		= &omap54xx_emif_hwmod_class,
 	.clkdm_name	= "emif_clkdm",
-	.flags		= HWMOD_INIT_NO_IDLE,
 	.main_clk	= "dpll_core_h11x2_ck",
 	.prcm = {
 		.omap4 = {
@@ -368,7 +367,6 @@ static struct omap_hwmod omap54xx_emif2_hwmod = {
 	.name		= "emif2",
 	.class		= &omap54xx_emif_hwmod_class,
 	.clkdm_name	= "emif_clkdm",
-	.flags		= HWMOD_INIT_NO_IDLE,
 	.main_clk	= "dpll_core_h11x2_ck",
 	.prcm = {
 		.omap4 = {
Signature · 1 line
1.7.9.5
#12 lore raw

PATCH v2 5/5 ARM: dts: AM335x-evmsk: Do not reset gpio0

In reply to Rajendra Nayak

Do not reset GPIO0 at boot-up because GPIO0 is used on AM335x EVM-SK to control VTT regulators on DDR3.

Without this EVM-SK boards fail to boot-up because of DDR3 corruption.

---
 arch/arm/boot/dts/am335x-evmsk.dts |    4 ++++
 1 file changed, 4 insertions(+)
arch/arm/boot/dts/am335x-evmsk.dts +4 -0
index 4f339fa..0e44d29 100644
@@ -419,3 +419,7 @@
 	phy_id = <&davinci_mdio>, <1>;
 	phy-mode = "rgmii-txid";
 };
+
+&gpio0 {
+	ti,no-reset-on-init;
+};
Signature · 1 line
1.7.9.5
#13 lore raw

PATCH 1/1 ARM: dts: omap4-panda-es: Do not reset gpio1

In reply to Rajendra Nayak

Do not reset GPIO1 at boot-up because GPIO 7 in GPIO1 block is used on OMAP4460 PandaBoard-ES to select voltage register in TPS62361 which supplies VDD_MPU.

Without this, OMAP4460 PandaBoard-ES boards fail to boot-up because MPU voltage switches over to VSET0 voltage value (boot voltage) which is not sufficient to operate the device at OPP100.

---
As explained here: http://marc.info/?l=u-boot&m=133066647800872&w=2

GPIO1 reset causes PandaBoard-ES to stop functioning. This depends on Patch series from Rajendra:
http://marc.info/?l=linux-doc&m=138131355520116&w=2
Applies on Benoit's for_13/dts branch:
https://git.kernel.org/cgit/linux/kernel/git/bcousson/linux-omap-dt.git/log/?h=for_3.13/dts

Tested on PandaBoard-ES without the u-boot uenv hack

 arch/arm/boot/dts/omap4-panda-es.dts |    4 ++++
 1 file changed, 4 insertions(+)
arch/arm/boot/dts/omap4-panda-es.dts +4 -0
index 56c4354..816d1c9 100644
@@ -62,3 +62,7 @@
 		gpios = <&gpio1 8 GPIO_ACTIVE_HIGH>;
 	};
 };
+
+&gpio1 {
+	 ti,no-reset-on-init;
+};
Signature · 1 line
1.7.9.5
maintainer
#14 lore raw

Re: [PATCH] ARM: dts: omap4-panda-es: Do not reset gpio1

Quoted 7 lines · Nishanth Menon

Do not reset GPIO1 at boot-up because GPIO 7 in GPIO1 block is used on OMAP4460 PandaBoard-ES to select voltage register in TPS62361 which supplies VDD_MPU.

Without this, OMAP4460 PandaBoard-ES boards fail to boot-up because MPU voltage switches over to VSET0 voltage value (boot voltage) which is not sufficient to operate the device at OPP100.

Is this needs as a fix for the -rc series?

Tony

Quoted 35 lines

---
As explained here: http://marc.info/?l=u-boot&m=133066647800872&w=2

GPIO1 reset causes PandaBoard-ES to stop functioning. This depends on Patch series from Rajendra:
http://marc.info/?l=linux-doc&m=138131355520116&w=2
Applies on Benoit's for_13/dts branch:
https://git.kernel.org/cgit/linux/kernel/git/bcousson/linux-omap-dt.git/log/?h=for_3.13/dts

Tested on PandaBoard-ES without the u-boot uenv hack

 arch/arm/boot/dts/omap4-panda-es.dts |    4 ++++
 1 file changed, 4 insertions(+)
arch/arm/boot/dts/omap4-panda-es.dts +4 -0
index 56c4354..816d1c9 100644
@@ -62,3 +62,7 @@
 		gpios = <&gpio1 8 GPIO_ACTIVE_HIGH>;
 	};
 };
+
+&gpio1 {
+	 ti,no-reset-on-init;
+};
Signature · 7 lines
1.7.9.5

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo at vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/
#15 lore raw

Re: [PATCH] ARM: dts: omap4-panda-es: Do not reset gpio1

Quoted 10 lines · Tony Lindgren
Quoted 7 lines · Nishanth Menon

Do not reset GPIO1 at boot-up because GPIO 7 in GPIO1 block is used on OMAP4460 PandaBoard-ES to select voltage register in TPS62361 which supplies VDD_MPU.

Without this, OMAP4460 PandaBoard-ES boards fail to boot-up because MPU voltage switches over to VSET0 voltage value (boot voltage) which is not sufficient to operate the device at OPP100.

Is this needs as a fix for the -rc series?

Nope, this is not a regression recently introduced, has been in since day one of Panda-ES support. Further, we still have decision pending on the dependent series from Rajendra[1] - will be good to see it go in along with AM335x-EVM.

[1] http://marc.info/?t=138131368300005&r=1&w=2

Signature · 2 lines
Regards,
Nishanth Menon
#16 lore raw

Re: [PATCH] ARM: dts: omap4-panda-es: Do not reset gpio1

In reply to Nishanth Menon

Hi Nishanth,

Quoted 9 lines · Nishanth Menon

Do not reset GPIO1 at boot-up because GPIO 7 in GPIO1 block is used on OMAP4460 PandaBoard-ES to select voltage register in TPS62361 which supplies VDD_MPU.

Without this, OMAP4460 PandaBoard-ES boards fail to boot-up because MPU voltage switches over to VSET0 voltage value (boot voltage) which is not sufficient to operate the device at OPP100.

Quoted 8 lines

---
As explained here: http://marc.info/?l=u-boot&m=133066647800872&w=2

GPIO1 reset causes PandaBoard-ES to stop functioning. This depends on Patch series from Rajendra:
http://marc.info/?l=linux-doc&m=138131355520116&w=2
Applies on Benoit's for_13/dts branch:
https://git.kernel.org/cgit/linux/kernel/git/bcousson/linux-omap-dt.git/log/?h=for_3.13/dts

I've just applied it after Rajendra's series.

Thanks,
Benoit

Quoted 19 lines

Tested on PandaBoard-ES without the u-boot uenv hack

  arch/arm/boot/dts/omap4-panda-es.dts |    4 ++++
  1 file changed, 4 insertions(+)
arch/arm/boot/dts/omap4-panda-es.dts +4 -0
index 56c4354..816d1c9 100644
@@ -62,3 +62,7 @@
  		gpios = <&gpio1 8 GPIO_ACTIVE_HIGH>;
  	};
  };
+
+&gpio1 {
+	 ti,no-reset-on-init;
+};
Signature · 4 lines
Beno?t Cousson
BayLibre
Embedded Linux Technology Lab
www.baylibre.com