Skip to contentSkip to conversation

ARM: OMAP5: Few updates discovered with es2.0

PATCH New Santosh Shilimkar 3 participants 19 messages Feb 20, 2013 – Mar 15, 2013

Subsystems (4)
lore
Share & tools
lore
#0 lore raw

cover ARM: OMAP5: Few updates discovered with es2.0

Here are few OMAP5 updates discovered during es2.0 silicon validation. Tested on OMAP5430 and OMAP5432 devices after applying out of tree datafiles for OMAP5.

Rajendra Nayak (1):
ARM: OMAP5: clock: No Freqsel on OMAP5 devices

Santosh Shilimkar (6):
ARM: OMAP5: Update SOC id detection code for ES2
ARM: OMAP5: timer: Update the clocksource name as per clock data
ARM: OMAP5: prm: Allow prm init to success on OMAP5 SOCs
ARM: OMAP5: Update SAR RAM base address
ARM: OMAP5: Update SAR memory layout for WakeupGen
ARM: OMAP5: Make errata i688 workaround available

Tero Kristo (1):
ARM: OMAP5: Reuse prm read_inst/write_inst

 arch/arm/mach-omap2/Kconfig            |    2 +-
 arch/arm/mach-omap2/dpll3xxx.c         |    6 ++++--
 arch/arm/mach-omap2/id.c               |   16 +++++++++++++---
 arch/arm/mach-omap2/io.c               |    9 +++++++++
 arch/arm/mach-omap2/omap4-common.c     |   10 ++++++++--
 arch/arm/mach-omap2/omap4-sar-layout.h |   14 +++++++-------
 arch/arm/mach-omap2/omap54xx.h         |    1 +
 arch/arm/mach-omap2/prm44xx.c          |    6 +++---
 arch/arm/mach-omap2/soc.h              |    2 ++
 arch/arm/mach-omap2/timer.c            |    5 +++--
 10 files changed, 51 insertions(+), 20 deletions(-)
Signature · 1 line
1.7.9.5
#1 lore raw

PATCH 1/8 ARM: OMAP5: Update SOC id detection code for ES2

Update OMAP5 ES2 idcode and make ES2 as default detection.

---
 arch/arm/mach-omap2/id.c  |   16 +++++++++++++---
 arch/arm/mach-omap2/soc.h |    2 ++
 2 files changed, 15 insertions(+), 3 deletions(-)
arch/arm/mach-omap2/id.c +13 -3
index 45cc7ed4..ed6f407 100644
@@ -519,22 +519,32 @@ void __init omap5xxx_check_revision(void)
 	case 0xb942:
 		switch (rev) {
 		case 0:
-		default:
 			omap_revision = OMAP5430_REV_ES1_0;
+			break;
+		case 1:
+			omap_revision = OMAP5430_REV_ES2_0;
+			break;
+		default:
+			omap_revision = OMAP5430_REV_ES2_0;
 		}
 		break;
 
 	case 0xb998:
 		switch (rev) {
 		case 0:
-		default:
 			omap_revision = OMAP5432_REV_ES1_0;
+			break;
+		case 1:
+			omap_revision = OMAP5432_REV_ES2_0;
+			break;
+		default:
+			omap_revision = OMAP5432_REV_ES2_0;
 		}
 		break;
 
 	default:
 		/* Unknown default to latest silicon rev as default*/
-		omap_revision = OMAP5430_REV_ES1_0;
+		omap_revision = OMAP5430_REV_ES2_0;
 	}
 
 	pr_info("OMAP%04x ES%d.0\n",
arch/arm/mach-omap2/soc.h +2 -0
index f31d907..4c90b3b 100644
@@ -404,7 +404,9 @@ IS_OMAP_TYPE(3430, 0x3430)
 
 #define OMAP54XX_CLASS		0x54000054
 #define OMAP5430_REV_ES1_0	(OMAP54XX_CLASS | (0x30 << 16) | (0x10 << 8))
+#define OMAP5430_REV_ES2_0	(OMAP54XX_CLASS | (0x30 << 16) | (0x20 << 8))
 #define OMAP5432_REV_ES1_0	(OMAP54XX_CLASS | (0x32 << 16) | (0x10 << 8))
+#define OMAP5432_REV_ES2_0	(OMAP54XX_CLASS | (0x32 << 16) | (0x20 << 8))
 
 void omap2xxx_check_revision(void);
 void omap3xxx_check_revision(void);
Signature · 1 line
1.7.9.5
#2 lore raw

PATCH 1/8 ARM: OMAP5: Update SOC id detection code for ES2

Hi,

arch/arm/mach-omap2/id.c void __init omap5xxx_check_revision(void) · line 541

Quoted 11 lines · Santosh Shilimkar
arch/arm/mach-omap2/id.c +6 -1
 	case 0xb998:
 		switch (rev) {
 		case 0:
-		default:
 			omap_revision = OMAP5432_REV_ES1_0;
+			break;
+		case 1:
+			omap_revision = OMAP5432_REV_ES2_0;
+			break;
+		default:
+			omap_revision = OMAP5432_REV_ES2_0;

no need for the duplication:

+ case 1:
+ default:
+ omap_revision = OMAP5432_REV_ES2_0;

will work just fine.

arch/arm/mach-omap2/id.c void __init omap5xxx_check_revision(void) · line 547

Quoted 7 lines
arch/arm/mach-omap2/id.c +1 -1
 		}
 		break;
 
 	default:
 		/* Unknown default to latest silicon rev as default*/
-		omap_revision = OMAP5430_REV_ES1_0;
+		omap_revision = OMAP5430_REV_ES2_0;

how about we default to 0xffffffff ? That's the highest possible version, just not sure if our revision macros can cope with it.

Signature · 8 lines
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/20130220/9913146e/attachment.sig>
#3 lore raw

PATCH 1/8 ARM: OMAP5: Update SOC id detection code for ES2

Quoted 21 lines · Felipe Balbi

Hi,

arch/arm/mach-omap2/id.c void __init omap5xxx_check_revision(void) · line 541

Quoted 11 lines · Santosh Shilimkar
  	case 0xb998:
  		switch (rev) {
  		case 0:
arch/arm/mach-omap2/id.c +6 -1
-		default:
  			omap_revision = OMAP5432_REV_ES1_0;
+			break;
+		case 1:
+			omap_revision = OMAP5432_REV_ES2_0;
+			break;
+		default:
+			omap_revision = OMAP5432_REV_ES2_0;

no need for the duplication:

arch/arm/mach-omap2/id.c +3 -0
+		case 1:
+		default:
+			omap_revision = OMAP5432_REV_ES2_0;
Quoted 1 line

will work just fine.

right. Will fix that.

Quoted 4 lines
Quoted 4 lines
  		}
  		break;
Quoted 8 lines

arch/arm/mach-omap2/id.c void __init omap5xxx_check_revision(void) · line 547

Quoted 4 lines

default:
/* Unknown default to latest silicon rev as default*/

arch/arm/mach-omap2/id.c +1 -1
-		omap_revision = OMAP5430_REV_ES1_0;
+		omap_revision = OMAP5430_REV_ES2_0;

how about we default to 0xffffffff ? That's the highest possible version, just not sure if our revision macros can cope with it.

:D
For record, it was decided that we will always have latest silicon spin as default and hence OMAP5_ES2

#4 lore raw

PATCH 1/8 ARM: OMAP5: Update SOC id detection code for ES2

Hi,

Quoted 11 lines · Santosh Shilimkar
Quoted 8 lines

arch/arm/mach-omap2/id.c void __init omap5xxx_check_revision(void) · line 547

Quoted 4 lines
arch/arm/mach-omap2/id.c +1 -1
 	default:
 		/* Unknown default to latest silicon rev as default*/
-		omap_revision = OMAP5430_REV_ES1_0;
+		omap_revision = OMAP5430_REV_ES2_0;

how about we default to 0xffffffff ? That's the highest possible version, just not sure if our revision macros can cope with it.

:D
For record, it was decided that we will always have latest silicon spin as default and hence OMAP5_ES2

but then this line will forever be modified. Every year or so :-p

Signature · 8 lines
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/20130220/3e95b62d/attachment.sig>
#5 lore raw

PATCH 1/8 ARM: OMAP5: Update SOC id detection code for ES2

Quoted 11 lines · Felipe Balbi

Hi,

Quoted 8 lines · Santosh Shilimkar
Quoted 8 lines

arch/arm/mach-omap2/id.c void __init omap5xxx_check_revision(void) · line 547

Quoted 4 lines

default:
/* Unknown default to latest silicon rev as default*/

arch/arm/mach-omap2/id.c +1 -1
-		omap_revision = OMAP5430_REV_ES1_0;
+		omap_revision = OMAP5430_REV_ES2_0;

how about we default to 0xffffffff ? That's the highest possible version, just not sure if our revision macros can cope with it.

Quoted 6 lines
Quoted 3 lines

:D
For record, it was decided that we will always have latest silicon spin as default and hence OMAP5_ES2

but then this line will forever be modified. Every year or so :-p

No... Unless and until we spin new silicon revision for OMAP5 every year. This is OMAP5 specific ID code function.

Regards,
Santosh

#6 lore raw

PATCH 1/8 ARM: OMAP5: Update SOC id detection code for ES2

Quoted 12 lines · Santosh Shilimkar
Quoted 11 lines · Felipe Balbi

Hi,

Quoted 8 lines · Santosh Shilimkar
Quoted 8 lines

arch/arm/mach-omap2/id.c void __init omap5xxx_check_revision(void) · line 547

Quoted 4 lines
arch/arm/mach-omap2/id.c +1 -1
 	default:
 		/* Unknown default to latest silicon rev as default*/
-		omap_revision = OMAP5430_REV_ES1_0;
+		omap_revision = OMAP5430_REV_ES2_0;

how about we default to 0xffffffff ? That's the highest possible version, just not sure if our revision macros can cope with it.

Quoted 8 lines
Quoted 6 lines
Quoted 3 lines

:D
For record, it was decided that we will always have latest silicon spin as default and hence OMAP5_ES2

but then this line will forever be modified. Every year or so :-p

No... Unless and until we spin new silicon revision for OMAP5 every year. This is OMAP5 specific ID code function.

oh, right. nevermind then

Signature · 8 lines
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/20130220/33ce4170/attachment.sig>
#7 lore raw

PATCH 2/8 ARM: OMAP5: timer: Update the clocksource name as per clock data

In reply to Santosh Shilimkar

OMAP5 clockdata has different sys clock clock node name. Fix the timer code to take care of it.

---
 arch/arm/mach-omap2/timer.c |    5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)
arch/arm/mach-omap2/timer.c +3 -2
index b8ad6e6..3c29450 100644
@@ -62,6 +62,7 @@
 #define OMAP2_MPU_SOURCE	"sys_ck"
 #define OMAP3_MPU_SOURCE	OMAP2_MPU_SOURCE
 #define OMAP4_MPU_SOURCE	"sys_clkin_ck"
+#define OMAP5_MPU_SOURCE	"sys_clkin"
 #define OMAP2_32K_SOURCE	"func_32k_ck"
 #define OMAP3_32K_SOURCE	"omap_32k_fck"
 #define OMAP4_32K_SOURCE	"sys_32k_ck"
@@ -494,7 +495,7 @@ static void __init realtime_counter_init(void)
 		pr_err("%s: ioremap failed\n", __func__);
 		return;
 	}
-	sys_clk = clk_get(NULL, "sys_clkin_ck");
+	sys_clk = clk_get(NULL, OMAP5_MPU_SOURCE);
 	if (IS_ERR(sys_clk)) {
 		pr_err("%s: failed to get system clock handle\n", __func__);
 		iounmap(base);
@@ -634,7 +635,7 @@ OMAP_SYS_TIMER(4, local);
 
 #ifdef CONFIG_SOC_OMAP5
 OMAP_SYS_32K_TIMER_INIT(5, 1, OMAP4_32K_SOURCE, "ti,timer-alwon",
-			2, OMAP4_MPU_SOURCE);
+			2, OMAP5_MPU_SOURCE);
 static void __init omap5_realtime_timer_init(void)
 {
 	int err;
Signature · 1 line
1.7.9.5
#8 lore raw

PATCH 2/8 ARM: OMAP5: timer: Update the clocksource name as per clock data

Hello.

Quoted 1 line · Santosh Shilimkar

OMAP5 clockdata has different sys clock clock node name. Fix the timer code

One "clock" too many. :-)

Quoted 3 lines

to take care of it.

WBR, Sergei

#9 lore raw

PATCH 2/8 ARM: OMAP5: timer: Update the clocksource name as per clock data

Quoted 9 lines · Sergei Shtylyov

Hello.

Quoted 2 lines · Santosh Shilimkar

OMAP5 clockdata has different sys clock clock node name. Fix the timer code

One "clock" too many. :-)

Indeed. Will fix that.

Regards
Santosh

#10 lore raw

PATCH 3/8 ARM: OMAP5: prm: Allow prm init to success on OMAP5 SOCs

In reply to Santosh Shilimkar

Allow prm init to success on OMAP5 SOCs.

---
 arch/arm/mach-omap2/prm44xx.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm/mach-omap2/prm44xx.c +1 -1
index c05a343..1aae198 100644
@@ -650,7 +650,7 @@ static struct prm_ll_data omap44xx_prm_ll_data = {
 
 int __init omap44xx_prm_init(void)
 {
-	if (!cpu_is_omap44xx())
+	if (!cpu_is_omap44xx() && !soc_is_omap54xx())
 		return 0;
 
 	return prm_register(&omap44xx_prm_ll_data);
Signature · 1 line
1.7.9.5
#11 lore raw

PATCH 3/8 ARM: OMAP5: prm: Allow prm init to success on OMAP5 SOCs

Quoted 1 line · Santosh Shilimkar

Allow prm init to success on OMAP5 SOCs.

s/success/succeed/

Signature · 8 lines
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/20130220/7ea35c30/attachment.sig>
#13 lore raw

PATCH 4/8 ARM: OMAP5: Reuse prm read_inst/write_inst

In reply to Santosh Shilimkar

From: Tero Kristo <[email protected]>

Make use of 'prm_base' so that prm read_inst/write_inst can work on OMAP5 devices.

---
 arch/arm/mach-omap2/prm44xx.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
arch/arm/mach-omap2/prm44xx.c +2 -2
index 1aae198..cb24b26 100644
@@ -81,13 +81,13 @@ static struct prm_reset_src_map omap44xx_prm_reset_src_map[] = {
 /* Read a register in a CM/PRM instance in the PRM module */
 u32 omap4_prm_read_inst_reg(s16 inst, u16 reg)
 {
-	return __raw_readl(OMAP44XX_PRM_REGADDR(inst, reg));
+	return __raw_readl(prm_base + inst + reg);
 }
 
 /* Write into a register in a CM/PRM instance in the PRM module */
 void omap4_prm_write_inst_reg(u32 val, s16 inst, u16 reg)
 {
-	__raw_writel(val, OMAP44XX_PRM_REGADDR(inst, reg));
+	__raw_writel(val, prm_base + inst + reg);
 }
 
 /* Read-modify-write a register in a PRM module. Caller must lock */
Signature · 1 line
1.7.9.5
#14 lore raw

PATCH 5/8 ARM: OMAP5: Update SAR RAM base address

In reply to Santosh Shilimkar

Update SAR RAM base address for OMAP5 based devices.

---
 arch/arm/mach-omap2/omap4-common.c |   10 ++++++++--
 arch/arm/mach-omap2/omap54xx.h     |    1 +
 2 files changed, 9 insertions(+), 2 deletions(-)
arch/arm/mach-omap2/omap4-common.c +8 -2
index 6897ae2..59399f8 100644
@@ -239,15 +239,21 @@ void __iomem *omap4_get_sar_ram_base(void)
  */
 static int __init omap4_sar_ram_init(void)
 {
+	unsigned long sar_base;
+
 	/*
 	 * To avoid code running on other OMAPs in
 	 * multi-omap builds
 	 */
-	if (!cpu_is_omap44xx())
+	if (cpu_is_omap44xx())
+		sar_base = OMAP44XX_SAR_RAM_BASE;
+	else if (soc_is_omap54xx())
+		sar_base = OMAP54XX_SAR_RAM_BASE;
+	else
 		return -ENOMEM;
 
 	/* Static mapping, never released */
-	sar_ram_base = ioremap(OMAP44XX_SAR_RAM_BASE, SZ_16K);
+	sar_ram_base = ioremap(sar_base, SZ_16K);
 	if (WARN_ON(!sar_ram_base))
 		return -ENOMEM;
 
arch/arm/mach-omap2/omap54xx.h +1 -0
index a2582bb..a086ba1 100644
@@ -28,5 +28,6 @@
 #define OMAP54XX_PRCM_MPU_BASE		0x48243000
 #define OMAP54XX_SCM_BASE		0x4a002000
 #define OMAP54XX_CTRL_BASE		0x4a002800
+#define OMAP54XX_SAR_RAM_BASE		0x4ae26000
 
 #endif /* __ASM_SOC_OMAP555554XX_H */
Signature · 1 line
1.7.9.5
#15 lore raw

PATCH 6/8 ARM: OMAP5: Update SAR memory layout for WakeupGen

In reply to Santosh Shilimkar

On OMAP5 es2 WakeupGen SAR register layout offset have changed. Update the layout accordingly.

---
 arch/arm/mach-omap2/omap4-sar-layout.h |   14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)
arch/arm/mach-omap2/omap4-sar-layout.h +7 -7
index e170fe8..9374175 100644
@@ -48,13 +48,13 @@
 #define SAR_BACKUP_STATUS_WAKEUPGEN		0x10
 
 /* WakeUpGen save restore offset from OMAP54XX_SAR_RAM_BASE */
-#define OMAP5_WAKEUPGENENB_OFFSET_CPU0		(SAR_BANK3_OFFSET + 0x8d4)
-#define OMAP5_WAKEUPGENENB_SECURE_OFFSET_CPU0	(SAR_BANK3_OFFSET + 0x8e8)
-#define OMAP5_WAKEUPGENENB_OFFSET_CPU1		(SAR_BANK3_OFFSET + 0x8fc)
-#define OMAP5_WAKEUPGENENB_SECURE_OFFSET_CPU1	(SAR_BANK3_OFFSET + 0x910)
-#define OMAP5_AUXCOREBOOT0_OFFSET		(SAR_BANK3_OFFSET + 0x924)
-#define OMAP5_AUXCOREBOOT1_OFFSET		(SAR_BANK3_OFFSET + 0x928)
-#define OMAP5_AMBA_IF_MODE_OFFSET		(SAR_BANK3_OFFSET + 0x92c)
+#define OMAP5_WAKEUPGENENB_OFFSET_CPU0		(SAR_BANK3_OFFSET + 0x9dc)
+#define OMAP5_WAKEUPGENENB_SECURE_OFFSET_CPU0	(SAR_BANK3_OFFSET + 0x9f0)
+#define OMAP5_WAKEUPGENENB_OFFSET_CPU1		(SAR_BANK3_OFFSET + 0xa04)
+#define OMAP5_WAKEUPGENENB_SECURE_OFFSET_CPU1	(SAR_BANK3_OFFSET + 0xa18)
+#define OMAP5_AUXCOREBOOT0_OFFSET		(SAR_BANK3_OFFSET + 0xa2c)
+#define OMAP5_AUXCOREBOOT1_OFFSET		(SAR_BANK3_OFFSET + 0x930)
+#define OMAP5_AMBA_IF_MODE_OFFSET		(SAR_BANK3_OFFSET + 0xa34)
 #define OMAP5_SAR_BACKUP_STATUS_OFFSET		(SAR_BANK3_OFFSET + 0x800)
 
 #endif
Signature · 1 line
1.7.9.5
#16 lore raw

PATCH 7/8 ARM: OMAP5: Make errata i688 workaround available

In reply to Santosh Shilimkar

Errata i688 is also applicable for OMAP5 based devices. Update the code so that it can be enabled on OMAP5 devices.

---
 arch/arm/mach-omap2/Kconfig |    2 +-
 arch/arm/mach-omap2/io.c    |    9 +++++++++
 2 files changed, 10 insertions(+), 1 deletion(-)
arch/arm/mach-omap2/Kconfig +1 -1
index 41b581f..63a12d4 100644
@@ -397,7 +397,7 @@ config OMAP3_SDRC_AC_TIMING
 
 config OMAP4_ERRATA_I688
 	bool "OMAP4 errata: Async Bridge Corruption"
-	depends on ARCH_OMAP4
+	depends on ARCH_OMAP4 || SOC_OMAP5
 	select ARCH_HAS_BARRIERS
 	help
 	  If a data is stalled inside asynchronous bridge because of back
arch/arm/mach-omap2/io.c +9 -0
index 2c3fdd6..2bef5a7 100644
@@ -271,6 +271,14 @@ static struct map_desc omap54xx_io_desc[] __initdata = {
 		.length		= L4_PER_54XX_SIZE,
 		.type		= MT_DEVICE,
 	},
+#ifdef CONFIG_OMAP4_ERRATA_I688
+	{
+		.virtual	= OMAP4_SRAM_VA,
+		.pfn		= __phys_to_pfn(OMAP4_SRAM_PA),
+		.length		= PAGE_SIZE,
+		.type		= MT_MEMORY_SO,
+	},
+#endif
 };
 #endif
 
@@ -323,6 +331,7 @@ void __init omap4_map_io(void)
 void __init omap5_map_io(void)
 {
 	iotable_init(omap54xx_io_desc, ARRAY_SIZE(omap54xx_io_desc));
+	omap_barriers_init();
 }
 #endif
 /*
Signature · 1 line
1.7.9.5
#17 lore raw

PATCH 8/8 ARM: OMAP5: clock: No Freqsel on OMAP5 devices

In reply to Santosh Shilimkar

From: Rajendra Nayak <[email protected]>

OMAP5 does not have freqsel either, so add the missing checks for !soc_is_omap54xx()

---
 arch/arm/mach-omap2/dpll3xxx.c |    6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
arch/arm/mach-omap2/dpll3xxx.c +4 -2
index 0a02aab5..6e3811b 100644
@@ -310,7 +310,8 @@ static int omap3_noncore_dpll_program(struct clk_hw_omap *clk, u16 freqsel)
 	 * Set jitter correction. No jitter correction for OMAP4 and 3630
 	 * since freqsel field is no longer present
 	 */
-	if (!soc_is_am33xx() && !cpu_is_omap44xx() && !cpu_is_omap3630()) {
+	if (!soc_is_am33xx() && !cpu_is_omap44xx() && !cpu_is_omap3630()
+            && !soc_is_omap54xx()) {
 		v = __raw_readl(dd->control_reg);
 		v &= ~dd->freqsel_mask;
 		v |= freqsel << __ffs(dd->freqsel_mask);
@@ -501,7 +502,8 @@ int omap3_noncore_dpll_set_rate(struct clk_hw *hw, unsigned long rate,
 			return -EINVAL;
 
 		/* No freqsel on OMAP4 and OMAP3630 */
-		if (!cpu_is_omap44xx() && !cpu_is_omap3630()) {
+		if (!cpu_is_omap44xx() && !cpu_is_omap3630()
+		    && !soc_is_omap54xx()) {
 			freqsel = _omap3_dpll_compute_freqsel(clk,
 						dd->last_rounded_n);
 			WARN_ON(!freqsel);
Signature · 1 line
1.7.9.5
#18 lore raw

PATCH 8/8 ARM: OMAP5: clock: No Freqsel on OMAP5 devices

Quoted 16 lines · Santosh Shilimkar

From: Rajendra Nayak <[email protected]>

OMAP5 does not have freqsel either, so add the missing checks for !soc_is_omap54xx()

---
 arch/arm/mach-omap2/dpll3xxx.c |    6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
arch/arm/mach-omap2/dpll3xxx.c +0 -0
index 0a02aab5..6e3811b 100644

arch/arm/mach-omap2/dpll3xxx.c static int omap3_noncore_dpll_program(struct clk_hw_omap *clk, u16 freqsel) · line 314

Quoted 6 lines
arch/arm/mach-omap2/dpll3xxx.c +2 -1
@@ -310,7 +310,8 @@ static int omap3_noncore_dpll_program(struct clk_hw_omap *clk, u16 freqsel)
 	 * Set jitter correction. No jitter correction for OMAP4 and 3630
 	 * since freqsel field is no longer present
 	 */
-	if (!soc_is_am33xx() && !cpu_is_omap44xx() && !cpu_is_omap3630()) {
+	if (!soc_is_am33xx() && !cpu_is_omap44xx() && !cpu_is_omap3630()
+            && !soc_is_omap54xx()) {

After re-looking at the freqsel patch, since the feature is available on only OMAP343X devices, its is better to just check that device to avoid future patching.

For the record, updated patch end of the email.

Regards,
Santosh

Quoted 1 line

From c31041957873bc524fe5d4fcd96688dc0d02e2db Mon Sep 17 00:00:00 2001

From: Rajendra Nayak <[email protected]>
Date: Tue, 29 Jan 2013 18:33:49 +0530
Subject: [PATCH 8/8] ARM: OMAP5: clock: No Freqsel on OMAP5 devices too

OMAP5 does not have freqsel either, so checks needs to be extended.

Infact only OMAP343X devices has the freqsel support, so fix the check accordingly so that future patching can be avoided.

---
 arch/arm/mach-omap2/dpll3xxx.c |   11 +++++------
 1 file changed, 5 insertions(+), 6 deletions(-)
arch/arm/mach-omap2/dpll3xxx.c +5 -6
index 3aed4b0..ae84c9d 100644
@@ -307,10 +307,10 @@ static int omap3_noncore_dpll_program(struct clk_hw_omap *clk, u16 freqsel)
 	_omap3_noncore_dpll_bypass(clk);
 
 	/*
-	 * Set jitter correction. No jitter correction for OMAP4 and 3630
-	 * since freqsel field is no longer present
+	 * Set jitter correction. Jitter correction applicable for OMAP343X
+	 * only since freqsel field is no longer present on other devices.
 	 */
-	if (!soc_is_am33xx() && !cpu_is_omap44xx() && !cpu_is_omap3630()) {
+	if (cpu_is_omap343x()) {
 		v = __raw_readl(dd->control_reg);
 		v &= ~dd->freqsel_mask;
 		v |= freqsel << __ffs(dd->freqsel_mask);
@@ -500,9 +500,8 @@ int omap3_noncore_dpll_set_rate(struct clk_hw *hw, unsigned long rate,
 		if (dd->last_rounded_rate == 0)
 			return -EINVAL;
 
-		/* No freqsel on AM335x, OMAP4 and OMAP3630 */
-		if (!soc_is_am33xx() && !cpu_is_omap44xx() &&
-		    !cpu_is_omap3630()) {
+		/* Freqsel is available only on OMAP343X devices */
+		if (cpu_is_omap343x()) {
 			freqsel = _omap3_dpll_compute_freqsel(clk,
 						dd->last_rounded_n);
 			WARN_ON(!freqsel);
Signature · 1 line
1.7.9.5