[U-Boot] [PATCH v03 0/2] OMAP3+: introduce generic Adaptive Body Bias Support

[U-Boot] [PATCH v03 0/2] OMAP3+: introduce generic Adaptive Body Bias Support

[Andrii Tseglytskyi]

Adaptive Body Biasing (ABB) modulates transistor bias voltages
dynamically in order to optimize switching speed versus leakage.
Adaptive Body-Bias ldos are present for some voltage domains:
starting with OMAP3630. There are three modes of operation:

* Bypass - the default, it just follows the vdd voltage
* Foward Body-Bias - applies voltage bias to increase transistor
  performance at the cost of power.  Used to operate safely at high
  OPPs.
* Reverse Body-Bias - applies voltage bias to decrease leakage and
  save power.  Used to save power at lower OPPs.

The application of FBB/RBB technique is determined for each unique
device in some process nodes, whereas, they are mandated on other
process nodes.

OMAP5 silicon requires ABB to be properly configured on early boot stage.
The following patch series adds generic ABB configuration and enables
it for OMAP5 silicons.

v02: series contains minor comments fixes.

v03: series contains:
 - clearing pending ABB interrupts before beginning of ABB configuration.
 - clearing ABB "setup" and "control" registers before beginning of
   ABB configuration.
 - removing superfluous wait on ABB "in transition" done status.

Andrii Tseglytskyi (2):
  OMAP3+: introduce generic ABB support
  OMAP5: add ABB setup for MPU voltage domain

 arch/arm/cpu/armv7/omap-common/Makefile        |    1 +
 arch/arm/cpu/armv7/omap-common/abb.c           |  137 ++++++++++++++++++++++++
 arch/arm/cpu/armv7/omap-common/clocks-common.c |    9 ++
 arch/arm/cpu/armv7/omap5/Makefile              |    1 +
 arch/arm/cpu/armv7/omap5/abb.c                 |   67 ++++++++++++
 arch/arm/cpu/armv7/omap5/prcm-regs.c           |    7 ++
 arch/arm/include/asm/arch-omap3/omap3.h        |    7 ++
 arch/arm/include/asm/arch-omap4/omap.h         |    7 ++
 arch/arm/include/asm/arch-omap5/omap.h         |   13 +++
 arch/arm/include/asm/omap_common.h             |   22 ++++
 10 files changed, 271 insertions(+)
 create mode 100644 arch/arm/cpu/armv7/omap-common/abb.c
 create mode 100644 arch/arm/cpu/armv7/omap5/abb.c

-- 
1.7.9.5

[U-Boot] [PATCH v03 1/2] OMAP3+: introduce generic ABB support

[Andrii Tseglytskyi]

Adaptive Body Biasing (ABB) modulates transistor bias voltages
dynamically in order to optimize switching speed versus leakage.
Adaptive Body-Bias ldos are present for some voltage domains
starting with OMAP3630. There are three modes of operation:

* Bypass - the default, it just follows the vdd voltage
* Foward Body-Bias - applies voltage bias to increase transistor
  performance at the cost of power.  Used to operate safely at high
  OPPs.
* Reverse Body-Bias - applies voltage bias to decrease leakage and
  save power.  Used to save power at lower OPPs.

Signed-off-by: Andrii Tseglytskyi <andrii.tseglytskyi@ti.com>
---
 arch/arm/cpu/armv7/omap-common/Makefile |    1 +
 arch/arm/cpu/armv7/omap-common/abb.c    |  137 +++++++++++++++++++++++++++++++
 arch/arm/cpu/armv7/omap5/Makefile       |    1 +
 arch/arm/cpu/armv7/omap5/abb.c          |   67 +++++++++++++++
 arch/arm/include/asm/arch-omap3/omap3.h |    7 ++
 arch/arm/include/asm/arch-omap4/omap.h  |    7 ++
 arch/arm/include/asm/arch-omap5/omap.h  |   13 +++
 arch/arm/include/asm/omap_common.h      |   22 +++++
 8 files changed, 255 insertions(+)
 create mode 100644 arch/arm/cpu/armv7/omap-common/abb.c
 create mode 100644 arch/arm/cpu/armv7/omap5/abb.c

diff --git a/arch/arm/cpu/armv7/omap-common/Makefile b/arch/arm/cpu/armv7/omap-common/Makefile
index 55e82ba..c4b9809 100644
--- a/arch/arm/cpu/armv7/omap-common/Makefile
+++ b/arch/arm/cpu/armv7/omap-common/Makefile
@@ -34,6 +34,7 @@ COBJS	+= hwinit-common.o
 COBJS	+= clocks-common.o
 COBJS	+= emif-common.o
 COBJS	+= vc.o
+COBJS	+= abb.o
 endif
 
 ifneq ($(CONFIG_AM33XX)$(CONFIG_OMAP44XX)$(CONFIG_OMAP54XX)$(CONFIG_TI814X),)
diff --git a/arch/arm/cpu/armv7/omap-common/abb.c b/arch/arm/cpu/armv7/omap-common/abb.c
new file mode 100644
index 0000000..87d1fb8
--- /dev/null
+++ b/arch/arm/cpu/armv7/omap-common/abb.c
@@ -0,0 +1,137 @@
+/*
+ *
+ * Adaptive Body Bias programming sequence for OMAP family
+ *
+ * (C) Copyright 2013
+ * Texas Instruments, <www.ti.com>
+ *
+ * Andrii Tseglytskyi <andrii.tseglytskyi@ti.com>
+ *
+ * See file CREDITS for list of people who contributed to this
+ * project.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of
+ * the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston,
+ * MA 02111-1307 USA
+ */
+
+#include <common.h>
+#include <asm/omap_common.h>
+#include <asm/io.h>
+#include <asm/arch/sys_proto.h>
+
+__weak s8 abb_setup_ldovbb(u32 fuse, u32 ldovbb)
+{
+	return -1;
+}
+
+static void abb_setup_timings(u32 setup)
+{
+	u32 sys_rate, sr2_cnt, clk_cycles;
+
+	/*
+	 * SR2_WTCNT_VALUE is the settling time for the ABB ldo after a
+	 * transition and must be programmed with the correct time at boot.
+	 * The value programmed into the register is the number of SYS_CLK
+	 * clock cycles that match a given wall time profiled for the ldo.
+	 * This value depends on:
+	 * settling time of ldo in micro-seconds (varies per OMAP family),
+	 * of clock cycles per SYS_CLK period (varies per OMAP family),
+	 * the SYS_CLK frequency in MHz (varies per board)
+	 * The formula is:
+	 *
+	 *		       ldo settling time (in micro-seconds)
+	 * SR2_WTCNT_VALUE = ------------------------------------------
+	 *		    (# system clock cycles) * (sys_clk period)
+	 *
+	 * Put another way:
+	 *
+	 * SR2_WTCNT_VALUE = settling time / (# SYS_CLK cycles / SYS_CLK rate))
+	 *
+	 * To avoid dividing by zero multiply both "# clock cycles" and
+	 * "settling time" by 10 such that the final result is the one we want.
+	 */
+
+	/* calculate SR2_WTCNT_VALUE */
+	sys_rate = DIV_ROUND(V_OSCK, 1000000);
+	clk_cycles = DIV_ROUND(OMAP_ABB_CLOCK_CYCLES * 10, sys_rate);
+	sr2_cnt = DIV_ROUND(OMAP_ABB_SETTLING_TIME * 10, clk_cycles);
+
+	setbits_le32(setup,
+		     sr2_cnt << (ffs(OMAP_ABB_SETUP_SR2_WTCNT_VALUE_MASK) - 1));
+}
+
+void abb_setup(u32 fuse, u32 ldovbb, u32 setup, u32 control,
+	       u32 txdone, u32 txdone_mask, u32 opp)
+{
+	u32 abb_type_mask, opp_sel_mask;
+
+	/* sanity check */
+	if (!setup || !control || !txdone)
+		return;
+
+	/* setup ABB only in case of Fast or Slow OPP */
+	switch (opp) {
+	case OMAP_ABB_FAST_OPP:
+		abb_type_mask = OMAP_ABB_SETUP_ACTIVE_FBB_SEL_MASK;
+		opp_sel_mask = OMAP_ABB_CONTROL_FAST_OPP_SEL_MASK;
+		break;
+	case OMAP_ABB_SLOW_OPP:
+		abb_type_mask = OMAP_ABB_SETUP_ACTIVE_RBB_SEL_MASK;
+		opp_sel_mask = OMAP_ABB_CONTROL_SLOW_OPP_SEL_MASK;
+		break;
+	default:
+	       return;
+	}
+
+	/*
+	 * For some OMAP silicons additional setup for LDOVBB register is
+	 * required. This is determined by data retrieved from corresponding
+	 * OPP EFUSE register. Data, which is retrieved from EFUSE - is
+	 * ABB enable/disable flag and VSET value, which must be copied
+	 * to LDOVBB register. If function call fails - return quietly,
+	 * it means no ABB is required for such silicon.
+	 *
+	 * For silicons, which don't require LDOVBB setup "fuse" and
+	 * "ldovbb" offsets are not defined. ABB will be initialized in
+	 * the common way for them.
+	 */
+	if (fuse && ldovbb) {
+		if (abb_setup_ldovbb(fuse, ldovbb))
+			return;
+	}
+
+	/* clear ABB registers */
+	writel(0, setup);
+	writel(0, control);
+
+	/* configure timings, based on oscillator value */
+	abb_setup_timings(setup);
+
+	/* clear pending interrupts before setup */
+	setbits_le32(txdone, txdone_mask);
+
+	/* select ABB type */
+	setbits_le32(setup, abb_type_mask | OMAP_ABB_SETUP_SR2EN_MASK);
+
+	/* initiate ABB ldo change */
+	setbits_le32(control, opp_sel_mask | OMAP_ABB_CONTROL_OPP_CHANGE_MASK);
+
+	/* wait until transition complete */
+	if (!wait_on_value(txdone_mask, txdone_mask, (void *)txdone, LDELAY))
+		puts("Error: ABB txdone is not set\n");
+
+	/* clear ABB tranxdone */
+	setbits_le32(txdone, txdone_mask);
+}
diff --git a/arch/arm/cpu/armv7/omap5/Makefile b/arch/arm/cpu/armv7/omap5/Makefile
index ce00e2c..6ff8dbb 100644
--- a/arch/arm/cpu/armv7/omap5/Makefile
+++ b/arch/arm/cpu/armv7/omap5/Makefile
@@ -30,6 +30,7 @@ COBJS	+= emif.o
 COBJS	+= sdram.o
 COBJS	+= prcm-regs.o
 COBJS	+= hw_data.o
+COBJS	+= abb.o
 
 SRCS	:= $(SOBJS:.o=.S) $(COBJS:.o=.c)
 OBJS	:= $(addprefix $(obj),$(COBJS) $(SOBJS))
diff --git a/arch/arm/cpu/armv7/omap5/abb.c b/arch/arm/cpu/armv7/omap5/abb.c
new file mode 100644
index 0000000..92470be
--- /dev/null
+++ b/arch/arm/cpu/armv7/omap5/abb.c
@@ -0,0 +1,67 @@
+/*
+ *
+ * Adaptive Body Bias programming sequence for OMAP5 family
+ *
+ * (C) Copyright 2013
+ * Texas Instruments, <www.ti.com>
+ *
+ * Andrii Tseglytskyi <andrii.tseglytskyi@ti.com>
+ *
+ * See file CREDITS for list of people who contributed to this
+ * project.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of
+ * the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston,
+ * MA 02111-1307 USA
+ */
+
+#include <common.h>
+#include <asm/omap_common.h>
+#include <asm/io.h>
+
+/*
+ * Setup LDOVBB for OMAP5.
+ * On OMAP5+ some ABB settings are fused. They are handled
+ * in the following way:
+ *
+ * 1. corresponding EFUSE register contains ABB enable bit
+ *    and VSET value
+ * 2. If ABB enable bit is set to 1, than ABB should be
+ *    enabled, otherwise ABB should be disabled
+ * 3. If ABB is enabled, than VSET value should be copied
+ *    to corresponding MUX control register
+ */
+s8 abb_setup_ldovbb(u32 fuse, u32 ldovbb)
+{
+	u32 vset;
+
+	/*
+	 * ABB parameters must be properly fused
+	 * otherwise ABB should be disabled
+	 */
+	vset = readl(fuse);
+	if (!(vset & OMAP5_ABB_FUSE_ENABLE_MASK))
+		return -1;
+
+	/* prepare VSET value for LDOVBB mux register */
+	vset &= OMAP5_ABB_FUSE_VSET_MASK;
+	vset >>= ffs(OMAP5_ABB_FUSE_VSET_MASK) - 1;
+	vset <<= ffs(OMAP5_ABB_LDOVBBMPU_VSET_OUT_MASK) - 1;
+	vset |= OMAP5_ABB_LDOVBBMPU_MUX_CTRL_MASK;
+
+	/* setup LDOVBB using fused value */
+	clrsetbits_le32(ldovbb,  OMAP5_ABB_LDOVBBMPU_VSET_OUT_MASK, vset);
+
+	return 0;
+}
diff --git a/arch/arm/include/asm/arch-omap3/omap3.h b/arch/arm/include/asm/arch-omap3/omap3.h
index 2b5e9ae..c57599a 100644
--- a/arch/arm/include/asm/arch-omap3/omap3.h
+++ b/arch/arm/include/asm/arch-omap3/omap3.h
@@ -253,4 +253,11 @@ struct gpio {
 
 #define OMAP3_EMU_HAL_START_HAL_CRITICAL	4
 
+/* ABB settings */
+#define OMAP_ABB_SETTLING_TIME		30
+#define OMAP_ABB_CLOCK_CYCLES		8
+
+/* ABB tranxdone mask */
+#define OMAP_ABB_MPU_TXDONE_MASK	(0x1 << 26)
+
 #endif
diff --git a/arch/arm/include/asm/arch-omap4/omap.h b/arch/arm/include/asm/arch-omap4/omap.h
index ad984da..559ad26 100644
--- a/arch/arm/include/asm/arch-omap4/omap.h
+++ b/arch/arm/include/asm/arch-omap4/omap.h
@@ -170,6 +170,13 @@ struct s32ktimer {
 #define CH_FLAGS_CHFLASH	(0x1 << 2)
 #define CH_FLAGS_CHMMCSD	(0x1 << 3)
 
+/* ABB settings */
+#define OMAP_ABB_SETTLING_TIME		50
+#define OMAP_ABB_CLOCK_CYCLES		16
+
+/* ABB tranxdone mask */
+#define OMAP_ABB_MPU_TXDONE_MASK	(0x1 << 7)
+
 #ifndef __ASSEMBLY__
 struct omap_boot_parameters {
 	char *boot_message;
diff --git a/arch/arm/include/asm/arch-omap5/omap.h b/arch/arm/include/asm/arch-omap5/omap.h
index 887fcaa..caa1234 100644
--- a/arch/arm/include/asm/arch-omap5/omap.h
+++ b/arch/arm/include/asm/arch-omap5/omap.h
@@ -243,6 +243,19 @@ struct s32ktimer {
 #define SRCODE_OVERRIDE_SEL_XS_SHIFT	0
 #define SRCODE_OVERRIDE_SEL_XS_MASK	(1 << 0)
 
+/* ABB settings */
+#define OMAP_ABB_SETTLING_TIME		50
+#define OMAP_ABB_CLOCK_CYCLES		16
+
+/* ABB tranxdone mask */
+#define OMAP_ABB_MPU_TXDONE_MASK		(0x1 << 7)
+
+/* ABB efuse masks */
+#define OMAP5_ABB_FUSE_VSET_MASK		(0x1F << 24)
+#define OMAP5_ABB_FUSE_ENABLE_MASK		(0x1 << 29)
+#define OMAP5_ABB_LDOVBBMPU_MUX_CTRL_MASK	(0x1 << 10)
+#define OMAP5_ABB_LDOVBBMPU_VSET_OUT_MASK	(0x1f << 0)
+
 #ifndef __ASSEMBLY__
 struct srcomp_params {
 	s8 divide_factor;
diff --git a/arch/arm/include/asm/omap_common.h b/arch/arm/include/asm/omap_common.h
index 091ddb5..4892c0a 100644
--- a/arch/arm/include/asm/omap_common.h
+++ b/arch/arm/include/asm/omap_common.h
@@ -240,6 +240,8 @@ struct prcm_regs {
 	u32 cm_l3init_fsusb_clkctrl;
 	u32 cm_l3init_ocp2scp1_clkctrl;
 
+	u32 prm_irqstatus_mpu_2;
+
 	/* cm2.l4per */
 	u32 cm_l4per_clkstctrl;
 	u32 cm_l4per_dynamicdep;
@@ -325,6 +327,8 @@ struct prcm_regs {
 	u32 prm_sldo_mpu_ctrl;
 	u32 prm_sldo_mm_setup;
 	u32 prm_sldo_mm_ctrl;
+	u32 prm_abbldo_mpu_setup;
+	u32 prm_abbldo_mpu_ctrl;
 
 	u32 cm_div_m4_dpll_core;
 	u32 cm_div_m5_dpll_core;
@@ -347,6 +351,7 @@ struct prcm_regs {
 
 struct omap_sys_ctrl_regs {
 	u32 control_status;
+	u32 control_std_fuse_opp_vdd_mpu_2;
 	u32 control_core_mmr_lock1;
 	u32 control_core_mmr_lock2;
 	u32 control_core_mmr_lock3;
@@ -416,6 +421,7 @@ struct omap_sys_ctrl_regs {
 	u32 control_port_emif2_sdram_config;
 	u32 control_emif1_sdram_config_ext;
 	u32 control_emif2_sdram_config_ext;
+	u32 control_wkup_ldovbb_mpu_voltage_ctrl;
 	u32 control_smart1nopmio_padconf_0;
 	u32 control_smart1nopmio_padconf_1;
 	u32 control_padconf_mode;
@@ -542,6 +548,9 @@ void enable_non_essential_clocks(void);
 void scale_vcores(struct vcores_data const *);
 u32 get_offset_code(u32 volt_offset, struct pmic_data *pmic);
 void do_scale_vcore(u32 vcore_reg, u32 volt_mv, struct pmic_data *pmic);
+void abb_setup(u32 fuse, u32 ldovbb, u32 setup, u32 control,
+	       u32 txdone, u32 txdone_mask, u32 opp);
+s8 abb_setup_ldovbb(u32 fuse, u32 ldovbb);
 
 /* Max value for DPLL multiplier M */
 #define OMAP_DPLL_MAX_N	127
@@ -552,6 +561,19 @@ void do_scale_vcore(u32 vcore_reg, u32 volt_mv, struct pmic_data *pmic);
 #define OMAP_INIT_CONTEXT_UBOOT_AFTER_SPL	2
 #define OMAP_INIT_CONTEXT_UBOOT_AFTER_CH	3
 
+/* ABB */
+#define OMAP_ABB_NOMINAL_OPP		0
+#define OMAP_ABB_FAST_OPP		1
+#define OMAP_ABB_SLOW_OPP		3
+#define OMAP_ABB_CONTROL_FAST_OPP_SEL_MASK		(0x1 << 0)
+#define OMAP_ABB_CONTROL_SLOW_OPP_SEL_MASK		(0x1 << 1)
+#define OMAP_ABB_CONTROL_OPP_CHANGE_MASK		(0x1 << 2)
+#define OMAP_ABB_CONTROL_SR2_IN_TRANSITION_MASK		(0x1 << 6)
+#define OMAP_ABB_SETUP_SR2EN_MASK			(0x1 << 0)
+#define OMAP_ABB_SETUP_ACTIVE_FBB_SEL_MASK		(0x1 << 2)
+#define OMAP_ABB_SETUP_ACTIVE_RBB_SEL_MASK		(0x1 << 1)
+#define OMAP_ABB_SETUP_SR2_WTCNT_VALUE_MASK		(0xff << 8)
+
 static inline u32 omap_revision(void)
 {
 	extern u32 *const omap_si_rev;
-- 
1.7.9.5

[U-Boot] [PATCH v03 2/2] OMAP5: add ABB setup for MPU voltage domain

[Andrii Tseglytskyi]

Patch adds a call of abb_setup() function, and proper registers
definitions needed for ABB setup sequence. ABB is initialized
for MPU voltage domain.

Signed-off-by: Andrii Tseglytskyi <andrii.tseglytskyi@ti.com>
---
 arch/arm/cpu/armv7/omap-common/clocks-common.c |    9 +++++++++
 arch/arm/cpu/armv7/omap5/prcm-regs.c           |    7 +++++++
 2 files changed, 16 insertions(+)

diff --git a/arch/arm/cpu/armv7/omap-common/clocks-common.c b/arch/arm/cpu/armv7/omap-common/clocks-common.c
index 2b955c7..bdecaf8 100644
--- a/arch/arm/cpu/armv7/omap-common/clocks-common.c
+++ b/arch/arm/cpu/armv7/omap-common/clocks-common.c
@@ -533,6 +533,15 @@ void scale_vcores(struct vcores_data const *vcores)
 	do_scale_vcore(vcores->mpu.addr, vcores->mpu.value,
 					  vcores->mpu.pmic);
 
+	/* Configure MPU ABB LDO after scale */
+	abb_setup((*ctrl)->control_std_fuse_opp_vdd_mpu_2,
+		  (*ctrl)->control_wkup_ldovbb_mpu_voltage_ctrl,
+		  (*prcm)->prm_abbldo_mpu_setup,
+		  (*prcm)->prm_abbldo_mpu_ctrl,
+		  (*prcm)->prm_irqstatus_mpu_2,
+		  OMAP_ABB_MPU_TXDONE_MASK,
+		  OMAP_ABB_FAST_OPP);
+
 	do_scale_vcore(vcores->mm.addr, vcores->mm.value,
 					  vcores->mm.pmic);
 
diff --git a/arch/arm/cpu/armv7/omap5/prcm-regs.c b/arch/arm/cpu/armv7/omap5/prcm-regs.c
index b8a61fe..b4b624e 100644
--- a/arch/arm/cpu/armv7/omap5/prcm-regs.c
+++ b/arch/arm/cpu/armv7/omap5/prcm-regs.c
@@ -311,6 +311,7 @@ struct prcm_regs const omap5_es1_prcm = {
 
 struct omap_sys_ctrl_regs const omap5_ctrl = {
 	.control_status				= 0x4A002134,
+	.control_std_fuse_opp_vdd_mpu_2		= 0x4A0021B4,
 	.control_paconf_global			= 0x4A002DA0,
 	.control_paconf_mode			= 0x4A002DA4,
 	.control_smart1io_padconf_0		= 0x4A002DA8,
@@ -358,6 +359,7 @@ struct omap_sys_ctrl_regs const omap5_ctrl = {
 	.control_port_emif2_sdram_config	= 0x4AE0C118,
 	.control_emif1_sdram_config_ext		= 0x4AE0C144,
 	.control_emif2_sdram_config_ext		= 0x4AE0C148,
+	.control_wkup_ldovbb_mpu_voltage_ctrl	= 0x4AE0C318,
 	.control_smart1nopmio_padconf_0		= 0x4AE0CDA0,
 	.control_smart1nopmio_padconf_1		= 0x4AE0CDA4,
 	.control_padconf_mode			= 0x4AE0CDA8,
@@ -709,6 +711,9 @@ struct prcm_regs const omap5_es2_prcm = {
 	.cm_l3init_fsusb_clkctrl = 0x4a0096d0,
 	.cm_l3init_ocp2scp1_clkctrl = 0x4a0096e0,
 
+	/* prm irqstatus regs */
+	.prm_irqstatus_mpu_2 = 0x4ae06014,
+
 	/* l4 wkup regs */
 	.cm_abe_pll_ref_clksel = 0x4ae0610c,
 	.cm_sys_clksel = 0x4ae06110,
@@ -739,6 +744,8 @@ struct prcm_regs const omap5_es2_prcm = {
 	.prm_sldo_mpu_ctrl = 0x4ae07cd0,
 	.prm_sldo_mm_setup = 0x4ae07cd4,
 	.prm_sldo_mm_ctrl = 0x4ae07cd8,
+	.prm_abbldo_mpu_setup = 0x4ae07cdc,
+	.prm_abbldo_mpu_ctrl = 0x4ae07ce0,
 };
 
 struct prcm_regs const dra7xx_prcm = {
-- 
1.7.9.5

[U-Boot] [PATCH v03 1/2] OMAP3+: introduce generic ABB support

[Nishanth Menon]

Hi Andrii,
We are almost there.. minor comments follow:
On 11:42-20130521, Andrii Tseglytskyi wrote:
[...]
> diff --git a/arch/arm/cpu/armv7/omap5/abb.c b/arch/arm/cpu/armv7/omap5/abb.c
> new file mode 100644
> index 0000000..92470be
> --- /dev/null
> +++ b/arch/arm/cpu/armv7/omap5/abb.c
> @@ -0,0 +1,67 @@
I might introduce this as part of patch #2... but no strong feelings
about it.
[...]
> +s8 abb_setup_ldovbb(u32 fuse, u32 ldovbb)
> +{
> +	u32 vset;
> +
> +	/*
> +	 * ABB parameters must be properly fused
> +	 * otherwise ABB should be disabled
> +	 */
> +	vset = readl(fuse);
> +	if (!(vset & OMAP5_ABB_FUSE_ENABLE_MASK))
> +		return -1;
> +
> +	/* prepare VSET value for LDOVBB mux register */
> +	vset &= OMAP5_ABB_FUSE_VSET_MASK;
> +	vset >>= ffs(OMAP5_ABB_FUSE_VSET_MASK) - 1;
> +	vset <<= ffs(OMAP5_ABB_LDOVBBMPU_VSET_OUT_MASK) - 1;
> +	vset |= OMAP5_ABB_LDOVBBMPU_MUX_CTRL_MASK;
> +
> +	/* setup LDOVBB using fused value */
> +	clrsetbits_le32(ldovbb,  OMAP5_ABB_LDOVBBMPU_VSET_OUT_MASK, vset);
OMAP5_ABB_LDOVBBMPU_MUX_CTRL_MASK wont get set :(
[...]

Other than this, I have no further comments.
-- 
Regards,
Nishanth Menon

[U-Boot] [PATCH v03 1/2] OMAP3+: introduce generic ABB support

[Andrii Tseglytskyi]

On 05/22/2013 11:56 PM, Nishanth Menon wrote:
> Hi Andrii,
> We are almost there.. minor comments follow:
> On 11:42-20130521, Andrii Tseglytskyi wrote:
> [...]
>> diff --git a/arch/arm/cpu/armv7/omap5/abb.c b/arch/arm/cpu/armv7/omap5/abb.c
>> new file mode 100644
>> index 0000000..92470be
>> --- /dev/null
>> +++ b/arch/arm/cpu/armv7/omap5/abb.c
>> @@ -0,0 +1,67 @@
> I might introduce this as part of patch #2... but no strong feelings
> about it.
> [...]
>> +s8 abb_setup_ldovbb(u32 fuse, u32 ldovbb)
>> +{
>> +	u32 vset;
>> +
>> +	/*
>> +	 * ABB parameters must be properly fused
>> +	 * otherwise ABB should be disabled
>> +	 */
>> +	vset = readl(fuse);
>> +	if (!(vset & OMAP5_ABB_FUSE_ENABLE_MASK))
>> +		return -1;
>> +
>> +	/* prepare VSET value for LDOVBB mux register */
>> +	vset &= OMAP5_ABB_FUSE_VSET_MASK;
>> +	vset >>= ffs(OMAP5_ABB_FUSE_VSET_MASK) - 1;
>> +	vset <<= ffs(OMAP5_ABB_LDOVBBMPU_VSET_OUT_MASK) - 1;
>> +	vset |= OMAP5_ABB_LDOVBBMPU_MUX_CTRL_MASK;
>> +
>> +	/* setup LDOVBB using fused value */
>> +	clrsetbits_le32(ldovbb,  OMAP5_ABB_LDOVBBMPU_VSET_OUT_MASK, vset);
> OMAP5_ABB_LDOVBBMPU_MUX_CTRL_MASK wont get set :(

Sorry didn't get... I have

vset |= *OMAP5_ABB_LDOVBBMPU_MUX_CTRL_MASK*;  <-- set here
clrsetbits_le32(ldovbb,  OMAP5_ABB_LDOVBBMPU_VSET_OUT_MASK, vset);


Maybe it would be more clear to rework to:

clrsetbits_le32(ldovbb,  OMAP5_ABB_LDOVBBMPU_VSET_OUT_MASK, vset | *OMAP5_ABB_LDOVBBMPU_MUX_CTRL_MASK*);


Is this your suggestion?

Regards,
Andrii
> [...]
>
> Other than this, I have no further comments.

[U-Boot] [PATCH v03 1/2] OMAP3+: introduce generic ABB support

[Andrii Tseglytskyi]

On 05/22/2013 11:56 PM, Nishanth Menon wrote:
> Hi Andrii,
> We are almost there.. minor comments follow:
> On 11:42-20130521, Andrii Tseglytskyi wrote:
> [...]
>> diff --git a/arch/arm/cpu/armv7/omap5/abb.c b/arch/arm/cpu/armv7/omap5/abb.c
>> new file mode 100644
>> index 0000000..92470be
>> --- /dev/null
>> +++ b/arch/arm/cpu/armv7/omap5/abb.c
>> @@ -0,0 +1,67 @@
> I might introduce this as part of patch #2... but no strong feelings
> about it.

I suggest to introduce OMAP5 logic in this patch. Now series is divided 
to 2 patches:

1. Control logic and needed definitions for all OMAPs -> so after this 
patch ABB can be used for OMAP3/4/5
2. ABB setup function call (in common code) and ABB registers add to 
OMAP5 data structures, so ABB will be configured for OMAP5 only in this 
case.

So, we have ->  1. complete control logic introduction + 2. using 
previously introduced logic.
I suggest to leave this as is.

> [...]
>> +s8 abb_setup_ldovbb(u32 fuse, u32 ldovbb)
>> +{
>> +	u32 vset;
>> +
>> +	/*
>> +	 * ABB parameters must be properly fused
>> +	 * otherwise ABB should be disabled
>> +	 */
>> +	vset = readl(fuse);
>> +	if (!(vset & OMAP5_ABB_FUSE_ENABLE_MASK))
>> +		return -1;
>> +
>> +	/* prepare VSET value for LDOVBB mux register */
>> +	vset &= OMAP5_ABB_FUSE_VSET_MASK;
>> +	vset >>= ffs(OMAP5_ABB_FUSE_VSET_MASK) - 1;
>> +	vset <<= ffs(OMAP5_ABB_LDOVBBMPU_VSET_OUT_MASK) - 1;
>> +	vset |= OMAP5_ABB_LDOVBBMPU_MUX_CTRL_MASK;
>> +
>> +	/* setup LDOVBB using fused value */
>> +	clrsetbits_le32(ldovbb,  OMAP5_ABB_LDOVBBMPU_VSET_OUT_MASK, vset);
> OMAP5_ABB_LDOVBBMPU_MUX_CTRL_MASK wont get set :(
> [...]
>
> Other than this, I have no further comments.

[U-Boot] [PATCH v03 1/2] OMAP3+: introduce generic ABB support

[Nishanth Menon]

On 10:39-20130523, Andrii Tseglytskyi wrote:
> On 05/22/2013 11:56 PM, Nishanth Menon wrote:
> >Hi Andrii,
> >We are almost there.. minor comments follow:
> >On 11:42-20130521, Andrii Tseglytskyi wrote:
> >[...]
> >>diff --git a/arch/arm/cpu/armv7/omap5/abb.c b/arch/arm/cpu/armv7/omap5/abb.c
> >>new file mode 100644
> >>index 0000000..92470be
> >>--- /dev/null
> >>+++ b/arch/arm/cpu/armv7/omap5/abb.c
> >>@@ -0,0 +1,67 @@
> >I might introduce this as part of patch #2... but no strong feelings
> >about it.
> >[...]
> >>+s8 abb_setup_ldovbb(u32 fuse, u32 ldovbb)
> >>+{
> >>+	u32 vset;
> >>+
> >>+	/*
> >>+	 * ABB parameters must be properly fused
> >>+	 * otherwise ABB should be disabled
> >>+	 */
> >>+	vset = readl(fuse);
> >>+	if (!(vset & OMAP5_ABB_FUSE_ENABLE_MASK))
> >>+		return -1;
> >>+
> >>+	/* prepare VSET value for LDOVBB mux register */
> >>+	vset &= OMAP5_ABB_FUSE_VSET_MASK;
> >>+	vset >>= ffs(OMAP5_ABB_FUSE_VSET_MASK) - 1;
> >>+	vset <<= ffs(OMAP5_ABB_LDOVBBMPU_VSET_OUT_MASK) - 1;
> >>+	vset |= OMAP5_ABB_LDOVBBMPU_MUX_CTRL_MASK;
> >>+
> >>+	/* setup LDOVBB using fused value */
> >>+	clrsetbits_le32(ldovbb,  OMAP5_ABB_LDOVBBMPU_VSET_OUT_MASK, vset);
> >OMAP5_ABB_LDOVBBMPU_MUX_CTRL_MASK wont get set :(
> 
> Sorry didn't get... I have
> 
> vset |= *OMAP5_ABB_LDOVBBMPU_MUX_CTRL_MASK*;  <-- set here
> clrsetbits_le32(ldovbb,  OMAP5_ABB_LDOVBBMPU_VSET_OUT_MASK, vset);
define clrsetbits(type, addr, clear, set) \
       out_##type((addr), (in_##type(addr) & ~(clear)) | (set))
Arrgh.. I missed this :(. I had confused this with *rmw (read modify
write) in Linux which would do set &= mask;

My Bad. the patch is doing the right thing.

In short,
Acked-by: Nishanth Menon <nm@ti.com>

-- 
Regards,
Nishanth Menon