[PATCH/RFC 0/4] CPUFreq for kzm9g

[PATCH/RFC 0/4] CPUFreq for kzm9g

[Guennadi Liakhovetski]

The first of these patches is a bug-fix, whereas the latter three patches 
are experimental and might want to be discussed. This patch set is based 
on an earlier Simon's topic/all+next branch snapshot. Attempts to use 
todays checkout failed - I couldn't even boot. Obviously, I'm also using 
my AS3711 patches and a number of other (not directly related) patches, 
that I'll be posting separately.

Guennadi Liakhovetski (4):
  ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy
  ARM: shmobile: sh73a0: add support for adjusting CPU frequency
  ARM: shmobile: sh73a0: run sh73a0_clock_init() only once
  ARM: shmobile: kzm9g-reference: add CPUFreq support

 arch/arm/boot/dts/sh73a0-kzm9g-reference.dts |   19 ++-
 arch/arm/mach-shmobile/Kconfig               |    2 +
 arch/arm/mach-shmobile/clock-sh73a0.c        |  224 +++++++++++++++++++++++++-
 3 files changed, 236 insertions(+), 9 deletions(-)

-- 
1.7.2.5

Thanks
Guennadi
---
Guennadi Liakhovetski, Ph.D.
Freelance Open-Source Software Developer
http://www.open-technology.de/

[PATCH/RFC 0/4] CPUFreq for kzm9g

[Guennadi Liakhovetski]

The first of these patches is a bug-fix, whereas the latter three patches 
are experimental and might want to be discussed. This patch set is based 
on an earlier Simon's topic/all+next branch snapshot. Attempts to use 
todays checkout failed - I couldn't even boot. Obviously, I'm also using 
my AS3711 patches and a number of other (not directly related) patches, 
that I'll be posting separately.

Guennadi Liakhovetski (4):
  ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy
  ARM: shmobile: sh73a0: add support for adjusting CPU frequency
  ARM: shmobile: sh73a0: run sh73a0_clock_init() only once
  ARM: shmobile: kzm9g-reference: add CPUFreq support

 arch/arm/boot/dts/sh73a0-kzm9g-reference.dts |   19 ++-
 arch/arm/mach-shmobile/Kconfig               |    2 +
 arch/arm/mach-shmobile/clock-sh73a0.c        |  224 +++++++++++++++++++++++++-
 3 files changed, 236 insertions(+), 9 deletions(-)

-- 
1.7.2.5

Thanks
Guennadi
---
Guennadi Liakhovetski, Ph.D.
Freelance Open-Source Software Developer
http://www.open-technology.de/

[PATCH/RFC 0/4] CPUFreq for kzm9g

[Guennadi Liakhovetski]

The first of these patches is a bug-fix, whereas the latter three patches 
are experimental and might want to be discussed. This patch set is based 
on an earlier Simon's topic/all+next branch snapshot. Attempts to use 
todays checkout failed - I couldn't even boot. Obviously, I'm also using 
my AS3711 patches and a number of other (not directly related) patches, 
that I'll be posting separately.

Guennadi Liakhovetski (4):
  ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy
  ARM: shmobile: sh73a0: add support for adjusting CPU frequency
  ARM: shmobile: sh73a0: run sh73a0_clock_init() only once
  ARM: shmobile: kzm9g-reference: add CPUFreq support

 arch/arm/boot/dts/sh73a0-kzm9g-reference.dts |   19 ++-
 arch/arm/mach-shmobile/Kconfig               |    2 +
 arch/arm/mach-shmobile/clock-sh73a0.c        |  224 +++++++++++++++++++++++++-
 3 files changed, 236 insertions(+), 9 deletions(-)

-- 
1.7.2.5

Thanks
Guennadi
---
Guennadi Liakhovetski, Ph.D.
Freelance Open-Source Software Developer
http://www.open-technology.de/

[PATCH 1/4] ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy

[Guennadi Liakhovetski]

Z and ZG clocks on sh73a0 have pll0 as their parent, not pll1.

Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
---
 arch/arm/mach-shmobile/clock-sh73a0.c |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/arm/mach-shmobile/clock-sh73a0.c b/arch/arm/mach-shmobile/clock-sh73a0.c
index 5fa106b..71843dd 100644
--- a/arch/arm/mach-shmobile/clock-sh73a0.c
+++ b/arch/arm/mach-shmobile/clock-sh73a0.c
@@ -265,12 +265,12 @@ enum { DIV4_I, DIV4_ZG, DIV4_M3, DIV4_B, DIV4_M1, DIV4_M2,
 
 static struct clk div4_clks[DIV4_NR] = {
 	[DIV4_I] = DIV4(FRQCRA, 20, 0xdff, CLK_ENABLE_ON_INIT),
-	[DIV4_ZG] = DIV4(FRQCRA, 16, 0xd7f, CLK_ENABLE_ON_INIT),
+	[DIV4_ZG] = SH_CLK_DIV4(&pll0_clk, FRQCRA, 16, 0xd7f, CLK_ENABLE_ON_INIT),
 	[DIV4_M3] = DIV4(FRQCRA, 12, 0x1dff, CLK_ENABLE_ON_INIT),
 	[DIV4_B] = DIV4(FRQCRA, 8, 0xdff, CLK_ENABLE_ON_INIT),
 	[DIV4_M1] = DIV4(FRQCRA, 4, 0x1dff, 0),
 	[DIV4_M2] = DIV4(FRQCRA, 0, 0x1dff, 0),
-	[DIV4_Z] = DIV4(FRQCRB, 24, 0x97f, 0),
+	[DIV4_Z] = SH_CLK_DIV4(&pll0_clk, FRQCRB, 24, 0x97f, 0),
 	[DIV4_ZTR] = DIV4(FRQCRB, 20, 0xdff, 0),
 	[DIV4_ZT] = DIV4(FRQCRB, 16, 0xdff, 0),
 	[DIV4_ZX] = DIV4(FRQCRB, 12, 0xdff, 0),
-- 
1.7.2.5

[PATCH 1/4] ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy

[Guennadi Liakhovetski]

Z and ZG clocks on sh73a0 have pll0 as their parent, not pll1.

Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
---
 arch/arm/mach-shmobile/clock-sh73a0.c |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/arm/mach-shmobile/clock-sh73a0.c b/arch/arm/mach-shmobile/clock-sh73a0.c
index 5fa106b..71843dd 100644
--- a/arch/arm/mach-shmobile/clock-sh73a0.c
+++ b/arch/arm/mach-shmobile/clock-sh73a0.c
@@ -265,12 +265,12 @@ enum { DIV4_I, DIV4_ZG, DIV4_M3, DIV4_B, DIV4_M1, DIV4_M2,
 
 static struct clk div4_clks[DIV4_NR] = {
 	[DIV4_I] = DIV4(FRQCRA, 20, 0xdff, CLK_ENABLE_ON_INIT),
-	[DIV4_ZG] = DIV4(FRQCRA, 16, 0xd7f, CLK_ENABLE_ON_INIT),
+	[DIV4_ZG] = SH_CLK_DIV4(&pll0_clk, FRQCRA, 16, 0xd7f, CLK_ENABLE_ON_INIT),
 	[DIV4_M3] = DIV4(FRQCRA, 12, 0x1dff, CLK_ENABLE_ON_INIT),
 	[DIV4_B] = DIV4(FRQCRA, 8, 0xdff, CLK_ENABLE_ON_INIT),
 	[DIV4_M1] = DIV4(FRQCRA, 4, 0x1dff, 0),
 	[DIV4_M2] = DIV4(FRQCRA, 0, 0x1dff, 0),
-	[DIV4_Z] = DIV4(FRQCRB, 24, 0x97f, 0),
+	[DIV4_Z] = SH_CLK_DIV4(&pll0_clk, FRQCRB, 24, 0x97f, 0),
 	[DIV4_ZTR] = DIV4(FRQCRB, 20, 0xdff, 0),
 	[DIV4_ZT] = DIV4(FRQCRB, 16, 0xdff, 0),
 	[DIV4_ZX] = DIV4(FRQCRB, 12, 0xdff, 0),
-- 
1.7.2.5

[PATCH 1/4] ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy

[Guennadi Liakhovetski]

Z and ZG clocks on sh73a0 have pll0 as their parent, not pll1.

Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
---
 arch/arm/mach-shmobile/clock-sh73a0.c |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/arm/mach-shmobile/clock-sh73a0.c b/arch/arm/mach-shmobile/clock-sh73a0.c
index 5fa106b..71843dd 100644
--- a/arch/arm/mach-shmobile/clock-sh73a0.c
+++ b/arch/arm/mach-shmobile/clock-sh73a0.c
@@ -265,12 +265,12 @@ enum { DIV4_I, DIV4_ZG, DIV4_M3, DIV4_B, DIV4_M1, DIV4_M2,
 
 static struct clk div4_clks[DIV4_NR] = {
 	[DIV4_I] = DIV4(FRQCRA, 20, 0xdff, CLK_ENABLE_ON_INIT),
-	[DIV4_ZG] = DIV4(FRQCRA, 16, 0xd7f, CLK_ENABLE_ON_INIT),
+	[DIV4_ZG] = SH_CLK_DIV4(&pll0_clk, FRQCRA, 16, 0xd7f, CLK_ENABLE_ON_INIT),
 	[DIV4_M3] = DIV4(FRQCRA, 12, 0x1dff, CLK_ENABLE_ON_INIT),
 	[DIV4_B] = DIV4(FRQCRA, 8, 0xdff, CLK_ENABLE_ON_INIT),
 	[DIV4_M1] = DIV4(FRQCRA, 4, 0x1dff, 0),
 	[DIV4_M2] = DIV4(FRQCRA, 0, 0x1dff, 0),
-	[DIV4_Z] = DIV4(FRQCRB, 24, 0x97f, 0),
+	[DIV4_Z] = SH_CLK_DIV4(&pll0_clk, FRQCRB, 24, 0x97f, 0),
 	[DIV4_ZTR] = DIV4(FRQCRB, 20, 0xdff, 0),
 	[DIV4_ZT] = DIV4(FRQCRB, 16, 0xdff, 0),
 	[DIV4_ZX] = DIV4(FRQCRB, 12, 0xdff, 0),
-- 
1.7.2.5

[PATCH/RFC 2/4] ARM: shmobile: sh73a0: add support for adjusting CPU frequency

[Guennadi Liakhovetski]

On SH73A0 the output of PLL0 is supplied to two dividers, feeding clock to
the CPU core and SGX. Lower CPU frequencies allow the use of lower supply
voltages and thus reduce power consumption.

Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
---

No, I don't like the idea of changing the parent frequency in the child 
clock driver very much either. But having a clock, that acts like this, 
allows the use of the generic cpufreq-cpu0 driver, which handles exactly 
one clock and one regulator. Instead of changing a physical clock driver 
to modify its parent, we could add a virtual clock, that would adjust them 
both. Otherwise, of course, we could write our own cpufreq driver, that 
would explicitly modify the 2 clocks, but that would be too hardware-
specific. I'm open for ideas.

 arch/arm/mach-shmobile/clock-sh73a0.c |  217 ++++++++++++++++++++++++++++++++-
 1 files changed, 213 insertions(+), 4 deletions(-)

diff --git a/arch/arm/mach-shmobile/clock-sh73a0.c b/arch/arm/mach-shmobile/clock-sh73a0.c
index 71843dd..3170482 100644
--- a/arch/arm/mach-shmobile/clock-sh73a0.c
+++ b/arch/arm/mach-shmobile/clock-sh73a0.c
@@ -16,6 +16,7 @@
  * along with this program; if not, write to the Free Software
  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  */
+#include <linux/delay.h>
 #include <linux/init.h>
 #include <linux/kernel.h>
 #include <linux/io.h>
@@ -156,18 +157,97 @@ static unsigned long pll_recalc(struct clk *clk)
 	return clk->parent->rate * mult;
 }
 
-static struct sh_clk_ops pll_clk_ops = {
+static int pll0_mult(struct clk *clk, unsigned long *rate)
+{
+	unsigned long mult, f_parent;
+
+	if (!clk->parent || !__clk_get(clk->parent))
+		return -ENODEV;
+
+	f_parent = clk_get_rate(clk->parent);
+	__clk_put(clk->parent);
+
+	if (WARN_ON(!f_parent))
+		/* Should never happen */
+		return -EINVAL;
+
+	mult = (*rate + f_parent / 2) / f_parent;
+
+	/* 27 <= multiplier <= 46 */
+	switch (mult) {
+	case 0 ... 27 / 2:
+		/* 1:1 - rate doesn't change */
+		return 1;
+	case 27 / 2 + 1 ... 27:
+		*rate *= 27;
+		return 27;
+	case 28 ... 45:
+		*rate *= mult;
+		return mult;
+	default:
+		*rate *= 46;
+		return 46;
+	}
+}
+
+static int pll0_set_rate(struct clk *clk, unsigned long rate)
+{
+	int mult = pll0_mult(clk, &rate), i;
+
+	if (mult < 0)
+		return mult;
+
+	if (mult = 1) {
+		/* 1:1 - switch PLL off */
+		__raw_writel(__raw_readl(PLLECR) & ~(1 << clk->enable_bit),
+			     PLLECR);
+		return 0;
+	}
+
+	i = __raw_readl(PLLECR) & (1 << (8 + clk->enable_bit));
+
+	__raw_writel((__raw_readl(clk->enable_reg) & ~(0x3f << 24)) |
+		     ((mult - 1) << 24), clk->enable_reg);
+
+	if (!i)
+		__raw_writel(__raw_readl(PLLECR) | (1 << clk->enable_bit),
+			     PLLECR);
+
+	for (i = 1000; i; i--)
+		if (__raw_readl(PLLECR) & (1 << (8 + clk->enable_bit)))
+			break;
+		else
+			cpu_relax();
+
+	return i ? 0 : -ETIMEDOUT;
+}
+
+static long pll0_round_rate(struct clk *clk, unsigned long rate)
+{
+	int ret = pll0_mult(clk, &rate);
+	if (ret < 0)
+		return ret;
+	return rate;
+}
+
+static struct sh_clk_ops pll0_clk_ops = {
 	.recalc		= pll_recalc,
+	.round_rate	= pll0_round_rate,
+	.set_rate	= pll0_set_rate,
 };
 
 static struct clk pll0_clk = {
-	.ops		= &pll_clk_ops,
+	.ops		= &pll0_clk_ops,
 	.flags		= CLK_ENABLE_ON_INIT,
 	.parent		= &main_clk,
 	.enable_reg	= (void __iomem *)PLL0CR,
 	.enable_bit	= 0,
 };
 
+static struct sh_clk_ops pll_clk_ops = {
+	.recalc		= pll_recalc,
+};
+
 static struct clk pll1_clk = {
 	.ops		= &pll_clk_ops,
 	.flags		= CLK_ENABLE_ON_INIT,
@@ -277,6 +357,126 @@ static struct clk div4_clks[DIV4_NR] = {
 	[DIV4_HP] = DIV4(FRQCRB, 4, 0xdff, 0),
 };
 
+static int (*div4_set_rate)(struct clk *clk, unsigned long rate);
+static unsigned long (*div4_recalc)(struct clk *clk);
+
+/* Supported system CPU (Z-clock) and PLL0 frequency combinations */
+static struct {
+	unsigned long zclk;
+	unsigned long pll0;
+} zclk_rate[] = {
+	{
+		.zclk = 1196000000,
+		.pll0 = 1196000000,
+	}, {
+		.zclk = 806000000,
+		.pll0 = 806000000,
+	}, {
+		.zclk = 403000000,
+		.pll0 = 806000000,
+	},
+};
+
+static int zclk_set_rate(struct clk *clk, unsigned long rate)
+{
+	int i, ret;
+	struct clk *pll0;
+
+	/* We only support frequencies from the zclk_rate table above */
+	for (i = 0; i < ARRAY_SIZE(zclk_rate); i++)
+		if (rate = zclk_rate[i].zclk)
+			break;
+
+	if (i = ARRAY_SIZE(zclk_rate)) {
+		pr_warning("%s(): unsupported CPU clock frequency %lu\n",
+			   __func__, rate);
+		return -EINVAL;
+	}
+
+	/* We could just use ->parent, but it's good to refcount */
+	pll0 = clk_get(NULL, "pll0_clk");
+	if (IS_ERR(pll0))
+		return PTR_ERR(pll0);
+
+	if (zclk_rate[i].pll0 != clk_get_rate(pll0)) {
+		/* cannot call clk_set_rate() - would cause a nested spinlock */
+		ret = pll0_set_rate(pll0, zclk_rate[i].pll0);
+		if (ret < 0)
+			goto esetrate;
+		pll0->rate = pll_recalc(pll0);
+		propagate_rate(pll0);
+	}
+
+	if (zclk_rate[i].pll0 = zclk_rate[i].zclk) {
+		/* 1:1 - switch off divider */
+		__raw_writel(__raw_readl(FRQCRB) & ~(1 << 28), FRQCRB);
+		ret = 0;
+	} else {
+		/* set the divider - call the DIV4 method */
+		ret = div4_set_rate(clk, rate);
+		if (ret < 0)
+			goto esetrate;
+
+		/* Enable the divider */
+		__raw_writel(__raw_readl(FRQCRB) | (1 << 28), FRQCRB);
+	}
+
+	/*
+	 * Kick the clock - this is also done in sh_clk_div_set_rate(), but we
+	 * want to check success
+	 */
+	__raw_writel(__raw_readl(FRQCRB) | (1 << 31), FRQCRB);
+	for (i = 1000; i; i--)
+		if (__raw_readl(FRQCRB) & (1 << 31))
+			cpu_relax();
+		else
+			break;
+	if (!i)
+		ret = -ETIMEDOUT;
+
+esetrate:
+	clk_put(pll0);
+	return ret;
+}
+
+static long zclk_round_rate(struct clk *clk, unsigned long rate)
+{
+	int i;
+
+	/* We only support frequencies from the zclk_rate table above */
+	for (i = 0; i < ARRAY_SIZE(zclk_rate); i++)
+		if (rate = zclk_rate[i].zclk)
+			break;
+
+	if (i = ARRAY_SIZE(zclk_rate)) {
+		pr_warning("%s(): unsupported CPU clock frequency %lu\n",
+			   __func__, rate);
+		return -EINVAL;
+	}
+
+	return rate;
+}
+
+static unsigned long zclk_recalc(struct clk *clk)
+{
+	/* Must recalculate frequencies, even if the divisor is unused ATM! */
+	unsigned long div_freq = div4_recalc(clk);
+
+	if (__raw_readl(FRQCRB) & (1 << 28))
+		return div_freq;
+
+	return clk_get_rate(clk->parent);
+}
+
+static void zclk_extend(void)
+{
+	div4_set_rate = div4_clks[DIV4_Z].ops->set_rate;
+	div4_recalc = div4_clks[DIV4_Z].ops->recalc;
+	div4_clks[DIV4_Z].ops->set_rate = zclk_set_rate;
+	div4_clks[DIV4_Z].ops->round_rate = zclk_round_rate;
+	div4_clks[DIV4_Z].ops->recalc = zclk_recalc;
+}
+
 enum { DIV6_VCK1, DIV6_VCK2, DIV6_VCK3, DIV6_ZB1,
 	DIV6_FLCTL, DIV6_SDHI0, DIV6_SDHI1, DIV6_SDHI2,
 	DIV6_FSIA, DIV6_FSIB, DIV6_SUB,
@@ -474,7 +674,7 @@ static struct clk *late_main_clks[] = {
 };
 
 enum { MSTP001,
-	MSTP129, MSTP128, MSTP127, MSTP126, MSTP125, MSTP118, MSTP116, MSTP100,
+	MSTP129, MSTP128, MSTP127, MSTP126, MSTP125, MSTP118, MSTP116, MSTP112, MSTP100,
 	MSTP219, MSTP218, MSTP217,
 	MSTP207, MSTP206, MSTP204, MSTP203, MSTP202, MSTP201, MSTP200,
 	MSTP331, MSTP329, MSTP328, MSTP325, MSTP323, MSTP322,
@@ -495,6 +695,7 @@ static struct clk mstp_clks[MSTP_NR] = {
 	[MSTP125] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR1, 25, 0), /* TMU0 */
 	[MSTP118] = MSTP(&div4_clks[DIV4_B], SMSTPCR1, 18, 0), /* DSITX0 */
 	[MSTP116] = MSTP(&div4_clks[DIV4_HP], SMSTPCR1, 16, 0), /* IIC0 */
+	[MSTP112] = MSTP(&div4_clks[DIV4_ZG], SMSTPCR1, 12, 0), /* SGX */
 	[MSTP100] = MSTP(&div4_clks[DIV4_B], SMSTPCR1, 0, 0), /* LCDC0 */
 	[MSTP219] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR2, 19, 0), /* SCIFA7 */
 	[MSTP218] = MSTP(&div4_clks[DIV4_HP], SMSTPCR2, 18, 0), /* SY-DMAC */
@@ -535,6 +736,10 @@ static struct clk mstp_clks[MSTP_NR] = {
 static struct clk_lookup lookups[] = {
 	/* main clocks */
 	CLKDEV_CON_ID("r_clk", &r_clk),
+	CLKDEV_CON_ID("pll0_clk", &pll0_clk),
+
+	/* DIV4 clocks */
+	CLKDEV_DEV_ID("cpu0", &div4_clks[DIV4_Z]), /* cpufreq-cpu0 */
 
 	/* DIV6 clocks */
 	CLKDEV_CON_ID("vck1_clk", &div6_clks[DIV6_VCK1]),
@@ -562,6 +767,7 @@ static struct clk_lookup lookups[] = {
 	CLKDEV_DEV_ID("sh-mipi-dsi.0", &mstp_clks[MSTP118]), /* DSITX */
 	CLKDEV_DEV_ID("i2c-sh_mobile.0", &mstp_clks[MSTP116]), /* I2C0 */
 	CLKDEV_DEV_ID("e6820000.i2c", &mstp_clks[MSTP116]), /* I2C0 */
+	CLKDEV_CON_ID("sgx_clk", &mstp_clks[MSTP112]),
 	CLKDEV_DEV_ID("sh_mobile_lcdc_fb.0", &mstp_clks[MSTP100]), /* LCDC0 */
 	CLKDEV_DEV_ID("sh-sci.7", &mstp_clks[MSTP219]), /* SCIFA7 */
 	CLKDEV_DEV_ID("sh-dma-engine.0", &mstp_clks[MSTP218]), /* SY-DMAC */
@@ -627,8 +833,11 @@ void __init sh73a0_clock_init(void)
 	for (k = 0; !ret && (k < ARRAY_SIZE(main_clks)); k++)
 		ret = clk_register(main_clks[k]);
 
-	if (!ret)
+	if (!ret) {
 		ret = sh_clk_div4_register(div4_clks, DIV4_NR, &div4_table);
+		if (!ret)
+			zclk_extend();
+	}
 
 	if (!ret)
 		ret = sh_clk_div6_reparent_register(div6_clks, DIV6_NR);
-- 
1.7.2.5

[PATCH/RFC 2/4] ARM: shmobile: sh73a0: add support for adjusting CPU frequency

[Guennadi Liakhovetski]

On SH73A0 the output of PLL0 is supplied to two dividers, feeding clock to
the CPU core and SGX. Lower CPU frequencies allow the use of lower supply
voltages and thus reduce power consumption.

Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
---

No, I don't like the idea of changing the parent frequency in the child 
clock driver very much either. But having a clock, that acts like this, 
allows the use of the generic cpufreq-cpu0 driver, which handles exactly 
one clock and one regulator. Instead of changing a physical clock driver 
to modify its parent, we could add a virtual clock, that would adjust them 
both. Otherwise, of course, we could write our own cpufreq driver, that 
would explicitly modify the 2 clocks, but that would be too hardware-
specific. I'm open for ideas.

 arch/arm/mach-shmobile/clock-sh73a0.c |  217 ++++++++++++++++++++++++++++++++-
 1 files changed, 213 insertions(+), 4 deletions(-)

diff --git a/arch/arm/mach-shmobile/clock-sh73a0.c b/arch/arm/mach-shmobile/clock-sh73a0.c
index 71843dd..3170482 100644
--- a/arch/arm/mach-shmobile/clock-sh73a0.c
+++ b/arch/arm/mach-shmobile/clock-sh73a0.c
@@ -16,6 +16,7 @@
  * along with this program; if not, write to the Free Software
  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  */
+#include <linux/delay.h>
 #include <linux/init.h>
 #include <linux/kernel.h>
 #include <linux/io.h>
@@ -156,18 +157,97 @@ static unsigned long pll_recalc(struct clk *clk)
 	return clk->parent->rate * mult;
 }
 
-static struct sh_clk_ops pll_clk_ops = {
+static int pll0_mult(struct clk *clk, unsigned long *rate)
+{
+	unsigned long mult, f_parent;
+
+	if (!clk->parent || !__clk_get(clk->parent))
+		return -ENODEV;
+
+	f_parent = clk_get_rate(clk->parent);
+	__clk_put(clk->parent);
+
+	if (WARN_ON(!f_parent))
+		/* Should never happen */
+		return -EINVAL;
+
+	mult = (*rate + f_parent / 2) / f_parent;
+
+	/* 27 <= multiplier <= 46 */
+	switch (mult) {
+	case 0 ... 27 / 2:
+		/* 1:1 - rate doesn't change */
+		return 1;
+	case 27 / 2 + 1 ... 27:
+		*rate *= 27;
+		return 27;
+	case 28 ... 45:
+		*rate *= mult;
+		return mult;
+	default:
+		*rate *= 46;
+		return 46;
+	}
+}
+
+static int pll0_set_rate(struct clk *clk, unsigned long rate)
+{
+	int mult = pll0_mult(clk, &rate), i;
+
+	if (mult < 0)
+		return mult;
+
+	if (mult == 1) {
+		/* 1:1 - switch PLL off */
+		__raw_writel(__raw_readl(PLLECR) & ~(1 << clk->enable_bit),
+			     PLLECR);
+		return 0;
+	}
+
+	i = __raw_readl(PLLECR) & (1 << (8 + clk->enable_bit));
+
+	__raw_writel((__raw_readl(clk->enable_reg) & ~(0x3f << 24)) |
+		     ((mult - 1) << 24), clk->enable_reg);
+
+	if (!i)
+		__raw_writel(__raw_readl(PLLECR) | (1 << clk->enable_bit),
+			     PLLECR);
+
+	for (i = 1000; i; i--)
+		if (__raw_readl(PLLECR) & (1 << (8 + clk->enable_bit)))
+			break;
+		else
+			cpu_relax();
+
+	return i ? 0 : -ETIMEDOUT;
+}
+
+static long pll0_round_rate(struct clk *clk, unsigned long rate)
+{
+	int ret = pll0_mult(clk, &rate);
+	if (ret < 0)
+		return ret;
+	return rate;
+}
+
+static struct sh_clk_ops pll0_clk_ops = {
 	.recalc		= pll_recalc,
+	.round_rate	= pll0_round_rate,
+	.set_rate	= pll0_set_rate,
 };
 
 static struct clk pll0_clk = {
-	.ops		= &pll_clk_ops,
+	.ops		= &pll0_clk_ops,
 	.flags		= CLK_ENABLE_ON_INIT,
 	.parent		= &main_clk,
 	.enable_reg	= (void __iomem *)PLL0CR,
 	.enable_bit	= 0,
 };
 
+static struct sh_clk_ops pll_clk_ops = {
+	.recalc		= pll_recalc,
+};
+
 static struct clk pll1_clk = {
 	.ops		= &pll_clk_ops,
 	.flags		= CLK_ENABLE_ON_INIT,
@@ -277,6 +357,126 @@ static struct clk div4_clks[DIV4_NR] = {
 	[DIV4_HP] = DIV4(FRQCRB, 4, 0xdff, 0),
 };
 
+static int (*div4_set_rate)(struct clk *clk, unsigned long rate);
+static unsigned long (*div4_recalc)(struct clk *clk);
+
+/* Supported system CPU (Z-clock) and PLL0 frequency combinations */
+static struct {
+	unsigned long zclk;
+	unsigned long pll0;
+} zclk_rate[] = {
+	{
+		.zclk = 1196000000,
+		.pll0 = 1196000000,
+	}, {
+		.zclk = 806000000,
+		.pll0 = 806000000,
+	}, {
+		.zclk = 403000000,
+		.pll0 = 806000000,
+	},
+};
+
+static int zclk_set_rate(struct clk *clk, unsigned long rate)
+{
+	int i, ret;
+	struct clk *pll0;
+
+	/* We only support frequencies from the zclk_rate table above */
+	for (i = 0; i < ARRAY_SIZE(zclk_rate); i++)
+		if (rate == zclk_rate[i].zclk)
+			break;
+
+	if (i == ARRAY_SIZE(zclk_rate)) {
+		pr_warning("%s(): unsupported CPU clock frequency %lu\n",
+			   __func__, rate);
+		return -EINVAL;
+	}
+
+	/* We could just use ->parent, but it's good to refcount */
+	pll0 = clk_get(NULL, "pll0_clk");
+	if (IS_ERR(pll0))
+		return PTR_ERR(pll0);
+
+	if (zclk_rate[i].pll0 != clk_get_rate(pll0)) {
+		/* cannot call clk_set_rate() - would cause a nested spinlock */
+		ret = pll0_set_rate(pll0, zclk_rate[i].pll0);
+		if (ret < 0)
+			goto esetrate;
+		pll0->rate = pll_recalc(pll0);
+		propagate_rate(pll0);
+	}
+
+	if (zclk_rate[i].pll0 == zclk_rate[i].zclk) {
+		/* 1:1 - switch off divider */
+		__raw_writel(__raw_readl(FRQCRB) & ~(1 << 28), FRQCRB);
+		ret = 0;
+	} else {
+		/* set the divider - call the DIV4 method */
+		ret = div4_set_rate(clk, rate);
+		if (ret < 0)
+			goto esetrate;
+
+		/* Enable the divider */
+		__raw_writel(__raw_readl(FRQCRB) | (1 << 28), FRQCRB);
+	}
+
+	/*
+	 * Kick the clock - this is also done in sh_clk_div_set_rate(), but we
+	 * want to check success
+	 */
+	__raw_writel(__raw_readl(FRQCRB) | (1 << 31), FRQCRB);
+	for (i = 1000; i; i--)
+		if (__raw_readl(FRQCRB) & (1 << 31))
+			cpu_relax();
+		else
+			break;
+	if (!i)
+		ret = -ETIMEDOUT;
+
+esetrate:
+	clk_put(pll0);
+	return ret;
+}
+
+static long zclk_round_rate(struct clk *clk, unsigned long rate)
+{
+	int i;
+
+	/* We only support frequencies from the zclk_rate table above */
+	for (i = 0; i < ARRAY_SIZE(zclk_rate); i++)
+		if (rate == zclk_rate[i].zclk)
+			break;
+
+	if (i == ARRAY_SIZE(zclk_rate)) {
+		pr_warning("%s(): unsupported CPU clock frequency %lu\n",
+			   __func__, rate);
+		return -EINVAL;
+	}
+
+	return rate;
+}
+
+static unsigned long zclk_recalc(struct clk *clk)
+{
+	/* Must recalculate frequencies, even if the divisor is unused ATM! */
+	unsigned long div_freq = div4_recalc(clk);
+
+	if (__raw_readl(FRQCRB) & (1 << 28))
+		return div_freq;
+
+	return clk_get_rate(clk->parent);
+}
+
+static void zclk_extend(void)
+{
+	div4_set_rate = div4_clks[DIV4_Z].ops->set_rate;
+	div4_recalc = div4_clks[DIV4_Z].ops->recalc;
+	div4_clks[DIV4_Z].ops->set_rate = zclk_set_rate;
+	div4_clks[DIV4_Z].ops->round_rate = zclk_round_rate;
+	div4_clks[DIV4_Z].ops->recalc = zclk_recalc;
+}
+
 enum { DIV6_VCK1, DIV6_VCK2, DIV6_VCK3, DIV6_ZB1,
 	DIV6_FLCTL, DIV6_SDHI0, DIV6_SDHI1, DIV6_SDHI2,
 	DIV6_FSIA, DIV6_FSIB, DIV6_SUB,
@@ -474,7 +674,7 @@ static struct clk *late_main_clks[] = {
 };
 
 enum { MSTP001,
-	MSTP129, MSTP128, MSTP127, MSTP126, MSTP125, MSTP118, MSTP116, MSTP100,
+	MSTP129, MSTP128, MSTP127, MSTP126, MSTP125, MSTP118, MSTP116, MSTP112, MSTP100,
 	MSTP219, MSTP218, MSTP217,
 	MSTP207, MSTP206, MSTP204, MSTP203, MSTP202, MSTP201, MSTP200,
 	MSTP331, MSTP329, MSTP328, MSTP325, MSTP323, MSTP322,
@@ -495,6 +695,7 @@ static struct clk mstp_clks[MSTP_NR] = {
 	[MSTP125] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR1, 25, 0), /* TMU0 */
 	[MSTP118] = MSTP(&div4_clks[DIV4_B], SMSTPCR1, 18, 0), /* DSITX0 */
 	[MSTP116] = MSTP(&div4_clks[DIV4_HP], SMSTPCR1, 16, 0), /* IIC0 */
+	[MSTP112] = MSTP(&div4_clks[DIV4_ZG], SMSTPCR1, 12, 0), /* SGX */
 	[MSTP100] = MSTP(&div4_clks[DIV4_B], SMSTPCR1, 0, 0), /* LCDC0 */
 	[MSTP219] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR2, 19, 0), /* SCIFA7 */
 	[MSTP218] = MSTP(&div4_clks[DIV4_HP], SMSTPCR2, 18, 0), /* SY-DMAC */
@@ -535,6 +736,10 @@ static struct clk mstp_clks[MSTP_NR] = {
 static struct clk_lookup lookups[] = {
 	/* main clocks */
 	CLKDEV_CON_ID("r_clk", &r_clk),
+	CLKDEV_CON_ID("pll0_clk", &pll0_clk),
+
+	/* DIV4 clocks */
+	CLKDEV_DEV_ID("cpu0", &div4_clks[DIV4_Z]), /* cpufreq-cpu0 */
 
 	/* DIV6 clocks */
 	CLKDEV_CON_ID("vck1_clk", &div6_clks[DIV6_VCK1]),
@@ -562,6 +767,7 @@ static struct clk_lookup lookups[] = {
 	CLKDEV_DEV_ID("sh-mipi-dsi.0", &mstp_clks[MSTP118]), /* DSITX */
 	CLKDEV_DEV_ID("i2c-sh_mobile.0", &mstp_clks[MSTP116]), /* I2C0 */
 	CLKDEV_DEV_ID("e6820000.i2c", &mstp_clks[MSTP116]), /* I2C0 */
+	CLKDEV_CON_ID("sgx_clk", &mstp_clks[MSTP112]),
 	CLKDEV_DEV_ID("sh_mobile_lcdc_fb.0", &mstp_clks[MSTP100]), /* LCDC0 */
 	CLKDEV_DEV_ID("sh-sci.7", &mstp_clks[MSTP219]), /* SCIFA7 */
 	CLKDEV_DEV_ID("sh-dma-engine.0", &mstp_clks[MSTP218]), /* SY-DMAC */
@@ -627,8 +833,11 @@ void __init sh73a0_clock_init(void)
 	for (k = 0; !ret && (k < ARRAY_SIZE(main_clks)); k++)
 		ret = clk_register(main_clks[k]);
 
-	if (!ret)
+	if (!ret) {
 		ret = sh_clk_div4_register(div4_clks, DIV4_NR, &div4_table);
+		if (!ret)
+			zclk_extend();
+	}
 
 	if (!ret)
 		ret = sh_clk_div6_reparent_register(div6_clks, DIV6_NR);
-- 
1.7.2.5

[PATCH/RFC 2/4] ARM: shmobile: sh73a0: add support for adjusting CPU frequency

[Guennadi Liakhovetski]

On SH73A0 the output of PLL0 is supplied to two dividers, feeding clock to
the CPU core and SGX. Lower CPU frequencies allow the use of lower supply
voltages and thus reduce power consumption.

Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
---

No, I don't like the idea of changing the parent frequency in the child 
clock driver very much either. But having a clock, that acts like this, 
allows the use of the generic cpufreq-cpu0 driver, which handles exactly 
one clock and one regulator. Instead of changing a physical clock driver 
to modify its parent, we could add a virtual clock, that would adjust them 
both. Otherwise, of course, we could write our own cpufreq driver, that 
would explicitly modify the 2 clocks, but that would be too hardware-
specific. I'm open for ideas.

 arch/arm/mach-shmobile/clock-sh73a0.c |  217 ++++++++++++++++++++++++++++++++-
 1 files changed, 213 insertions(+), 4 deletions(-)

diff --git a/arch/arm/mach-shmobile/clock-sh73a0.c b/arch/arm/mach-shmobile/clock-sh73a0.c
index 71843dd..3170482 100644
--- a/arch/arm/mach-shmobile/clock-sh73a0.c
+++ b/arch/arm/mach-shmobile/clock-sh73a0.c
@@ -16,6 +16,7 @@
  * along with this program; if not, write to the Free Software
  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  */
+#include <linux/delay.h>
 #include <linux/init.h>
 #include <linux/kernel.h>
 #include <linux/io.h>
@@ -156,18 +157,97 @@ static unsigned long pll_recalc(struct clk *clk)
 	return clk->parent->rate * mult;
 }
 
-static struct sh_clk_ops pll_clk_ops = {
+static int pll0_mult(struct clk *clk, unsigned long *rate)
+{
+	unsigned long mult, f_parent;
+
+	if (!clk->parent || !__clk_get(clk->parent))
+		return -ENODEV;
+
+	f_parent = clk_get_rate(clk->parent);
+	__clk_put(clk->parent);
+
+	if (WARN_ON(!f_parent))
+		/* Should never happen */
+		return -EINVAL;
+
+	mult = (*rate + f_parent / 2) / f_parent;
+
+	/* 27 <= multiplier <= 46 */
+	switch (mult) {
+	case 0 ... 27 / 2:
+		/* 1:1 - rate doesn't change */
+		return 1;
+	case 27 / 2 + 1 ... 27:
+		*rate *= 27;
+		return 27;
+	case 28 ... 45:
+		*rate *= mult;
+		return mult;
+	default:
+		*rate *= 46;
+		return 46;
+	}
+}
+
+static int pll0_set_rate(struct clk *clk, unsigned long rate)
+{
+	int mult = pll0_mult(clk, &rate), i;
+
+	if (mult < 0)
+		return mult;
+
+	if (mult == 1) {
+		/* 1:1 - switch PLL off */
+		__raw_writel(__raw_readl(PLLECR) & ~(1 << clk->enable_bit),
+			     PLLECR);
+		return 0;
+	}
+
+	i = __raw_readl(PLLECR) & (1 << (8 + clk->enable_bit));
+
+	__raw_writel((__raw_readl(clk->enable_reg) & ~(0x3f << 24)) |
+		     ((mult - 1) << 24), clk->enable_reg);
+
+	if (!i)
+		__raw_writel(__raw_readl(PLLECR) | (1 << clk->enable_bit),
+			     PLLECR);
+
+	for (i = 1000; i; i--)
+		if (__raw_readl(PLLECR) & (1 << (8 + clk->enable_bit)))
+			break;
+		else
+			cpu_relax();
+
+	return i ? 0 : -ETIMEDOUT;
+}
+
+static long pll0_round_rate(struct clk *clk, unsigned long rate)
+{
+	int ret = pll0_mult(clk, &rate);
+	if (ret < 0)
+		return ret;
+	return rate;
+}
+
+static struct sh_clk_ops pll0_clk_ops = {
 	.recalc		= pll_recalc,
+	.round_rate	= pll0_round_rate,
+	.set_rate	= pll0_set_rate,
 };
 
 static struct clk pll0_clk = {
-	.ops		= &pll_clk_ops,
+	.ops		= &pll0_clk_ops,
 	.flags		= CLK_ENABLE_ON_INIT,
 	.parent		= &main_clk,
 	.enable_reg	= (void __iomem *)PLL0CR,
 	.enable_bit	= 0,
 };
 
+static struct sh_clk_ops pll_clk_ops = {
+	.recalc		= pll_recalc,
+};
+
 static struct clk pll1_clk = {
 	.ops		= &pll_clk_ops,
 	.flags		= CLK_ENABLE_ON_INIT,
@@ -277,6 +357,126 @@ static struct clk div4_clks[DIV4_NR] = {
 	[DIV4_HP] = DIV4(FRQCRB, 4, 0xdff, 0),
 };
 
+static int (*div4_set_rate)(struct clk *clk, unsigned long rate);
+static unsigned long (*div4_recalc)(struct clk *clk);
+
+/* Supported system CPU (Z-clock) and PLL0 frequency combinations */
+static struct {
+	unsigned long zclk;
+	unsigned long pll0;
+} zclk_rate[] = {
+	{
+		.zclk = 1196000000,
+		.pll0 = 1196000000,
+	}, {
+		.zclk = 806000000,
+		.pll0 = 806000000,
+	}, {
+		.zclk = 403000000,
+		.pll0 = 806000000,
+	},
+};
+
+static int zclk_set_rate(struct clk *clk, unsigned long rate)
+{
+	int i, ret;
+	struct clk *pll0;
+
+	/* We only support frequencies from the zclk_rate table above */
+	for (i = 0; i < ARRAY_SIZE(zclk_rate); i++)
+		if (rate == zclk_rate[i].zclk)
+			break;
+
+	if (i == ARRAY_SIZE(zclk_rate)) {
+		pr_warning("%s(): unsupported CPU clock frequency %lu\n",
+			   __func__, rate);
+		return -EINVAL;
+	}
+
+	/* We could just use ->parent, but it's good to refcount */
+	pll0 = clk_get(NULL, "pll0_clk");
+	if (IS_ERR(pll0))
+		return PTR_ERR(pll0);
+
+	if (zclk_rate[i].pll0 != clk_get_rate(pll0)) {
+		/* cannot call clk_set_rate() - would cause a nested spinlock */
+		ret = pll0_set_rate(pll0, zclk_rate[i].pll0);
+		if (ret < 0)
+			goto esetrate;
+		pll0->rate = pll_recalc(pll0);
+		propagate_rate(pll0);
+	}
+
+	if (zclk_rate[i].pll0 == zclk_rate[i].zclk) {
+		/* 1:1 - switch off divider */
+		__raw_writel(__raw_readl(FRQCRB) & ~(1 << 28), FRQCRB);
+		ret = 0;
+	} else {
+		/* set the divider - call the DIV4 method */
+		ret = div4_set_rate(clk, rate);
+		if (ret < 0)
+			goto esetrate;
+
+		/* Enable the divider */
+		__raw_writel(__raw_readl(FRQCRB) | (1 << 28), FRQCRB);
+	}
+
+	/*
+	 * Kick the clock - this is also done in sh_clk_div_set_rate(), but we
+	 * want to check success
+	 */
+	__raw_writel(__raw_readl(FRQCRB) | (1 << 31), FRQCRB);
+	for (i = 1000; i; i--)
+		if (__raw_readl(FRQCRB) & (1 << 31))
+			cpu_relax();
+		else
+			break;
+	if (!i)
+		ret = -ETIMEDOUT;
+
+esetrate:
+	clk_put(pll0);
+	return ret;
+}
+
+static long zclk_round_rate(struct clk *clk, unsigned long rate)
+{
+	int i;
+
+	/* We only support frequencies from the zclk_rate table above */
+	for (i = 0; i < ARRAY_SIZE(zclk_rate); i++)
+		if (rate == zclk_rate[i].zclk)
+			break;
+
+	if (i == ARRAY_SIZE(zclk_rate)) {
+		pr_warning("%s(): unsupported CPU clock frequency %lu\n",
+			   __func__, rate);
+		return -EINVAL;
+	}
+
+	return rate;
+}
+
+static unsigned long zclk_recalc(struct clk *clk)
+{
+	/* Must recalculate frequencies, even if the divisor is unused ATM! */
+	unsigned long div_freq = div4_recalc(clk);
+
+	if (__raw_readl(FRQCRB) & (1 << 28))
+		return div_freq;
+
+	return clk_get_rate(clk->parent);
+}
+
+static void zclk_extend(void)
+{
+	div4_set_rate = div4_clks[DIV4_Z].ops->set_rate;
+	div4_recalc = div4_clks[DIV4_Z].ops->recalc;
+	div4_clks[DIV4_Z].ops->set_rate = zclk_set_rate;
+	div4_clks[DIV4_Z].ops->round_rate = zclk_round_rate;
+	div4_clks[DIV4_Z].ops->recalc = zclk_recalc;
+}
+
 enum { DIV6_VCK1, DIV6_VCK2, DIV6_VCK3, DIV6_ZB1,
 	DIV6_FLCTL, DIV6_SDHI0, DIV6_SDHI1, DIV6_SDHI2,
 	DIV6_FSIA, DIV6_FSIB, DIV6_SUB,
@@ -474,7 +674,7 @@ static struct clk *late_main_clks[] = {
 };
 
 enum { MSTP001,
-	MSTP129, MSTP128, MSTP127, MSTP126, MSTP125, MSTP118, MSTP116, MSTP100,
+	MSTP129, MSTP128, MSTP127, MSTP126, MSTP125, MSTP118, MSTP116, MSTP112, MSTP100,
 	MSTP219, MSTP218, MSTP217,
 	MSTP207, MSTP206, MSTP204, MSTP203, MSTP202, MSTP201, MSTP200,
 	MSTP331, MSTP329, MSTP328, MSTP325, MSTP323, MSTP322,
@@ -495,6 +695,7 @@ static struct clk mstp_clks[MSTP_NR] = {
 	[MSTP125] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR1, 25, 0), /* TMU0 */
 	[MSTP118] = MSTP(&div4_clks[DIV4_B], SMSTPCR1, 18, 0), /* DSITX0 */
 	[MSTP116] = MSTP(&div4_clks[DIV4_HP], SMSTPCR1, 16, 0), /* IIC0 */
+	[MSTP112] = MSTP(&div4_clks[DIV4_ZG], SMSTPCR1, 12, 0), /* SGX */
 	[MSTP100] = MSTP(&div4_clks[DIV4_B], SMSTPCR1, 0, 0), /* LCDC0 */
 	[MSTP219] = MSTP(&div6_clks[DIV6_SUB], SMSTPCR2, 19, 0), /* SCIFA7 */
 	[MSTP218] = MSTP(&div4_clks[DIV4_HP], SMSTPCR2, 18, 0), /* SY-DMAC */
@@ -535,6 +736,10 @@ static struct clk mstp_clks[MSTP_NR] = {
 static struct clk_lookup lookups[] = {
 	/* main clocks */
 	CLKDEV_CON_ID("r_clk", &r_clk),
+	CLKDEV_CON_ID("pll0_clk", &pll0_clk),
+
+	/* DIV4 clocks */
+	CLKDEV_DEV_ID("cpu0", &div4_clks[DIV4_Z]), /* cpufreq-cpu0 */
 
 	/* DIV6 clocks */
 	CLKDEV_CON_ID("vck1_clk", &div6_clks[DIV6_VCK1]),
@@ -562,6 +767,7 @@ static struct clk_lookup lookups[] = {
 	CLKDEV_DEV_ID("sh-mipi-dsi.0", &mstp_clks[MSTP118]), /* DSITX */
 	CLKDEV_DEV_ID("i2c-sh_mobile.0", &mstp_clks[MSTP116]), /* I2C0 */
 	CLKDEV_DEV_ID("e6820000.i2c", &mstp_clks[MSTP116]), /* I2C0 */
+	CLKDEV_CON_ID("sgx_clk", &mstp_clks[MSTP112]),
 	CLKDEV_DEV_ID("sh_mobile_lcdc_fb.0", &mstp_clks[MSTP100]), /* LCDC0 */
 	CLKDEV_DEV_ID("sh-sci.7", &mstp_clks[MSTP219]), /* SCIFA7 */
 	CLKDEV_DEV_ID("sh-dma-engine.0", &mstp_clks[MSTP218]), /* SY-DMAC */
@@ -627,8 +833,11 @@ void __init sh73a0_clock_init(void)
 	for (k = 0; !ret && (k < ARRAY_SIZE(main_clks)); k++)
 		ret = clk_register(main_clks[k]);
 
-	if (!ret)
+	if (!ret) {
 		ret = sh_clk_div4_register(div4_clks, DIV4_NR, &div4_table);
+		if (!ret)
+			zclk_extend();
+	}
 
 	if (!ret)
 		ret = sh_clk_div6_reparent_register(div6_clks, DIV6_NR);
-- 
1.7.2.5

[PATCH/RFC 3/4] ARM: shmobile: sh73a0: run sh73a0_clock_init() only once

[Guennadi Liakhovetski]

sh73a0_clock_init() is called from sh73a0_earlytimer_init() and
sh73a0_add_standard_devices_dt(), take care to only run it once.

Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
---

This patch shouldn't be needed, sh73a0_clock_init() shouldn't be called 
twice. It's here only as an illustration, how I actually was testing.

 arch/arm/mach-shmobile/clock-sh73a0.c |    3 +++
 1 files changed, 3 insertions(+), 0 deletions(-)

diff --git a/arch/arm/mach-shmobile/clock-sh73a0.c b/arch/arm/mach-shmobile/clock-sh73a0.c
index 3170482..f9a015d 100644
--- a/arch/arm/mach-shmobile/clock-sh73a0.c
+++ b/arch/arm/mach-shmobile/clock-sh73a0.c
@@ -809,6 +809,9 @@ void __init sh73a0_clock_init(void)
 {
 	int k, ret = 0;
 
+	if (main_clk.parent)
+		return;
+
 	/* Set SDHI clocks to a known state */
 	__raw_writel(0x108, SD0CKCR);
 	__raw_writel(0x108, SD1CKCR);
-- 
1.7.2.5

[PATCH/RFC 3/4] ARM: shmobile: sh73a0: run sh73a0_clock_init() only once

[Guennadi Liakhovetski]

sh73a0_clock_init() is called from sh73a0_earlytimer_init() and
sh73a0_add_standard_devices_dt(), take care to only run it once.

Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
---

This patch shouldn't be needed, sh73a0_clock_init() shouldn't be called 
twice. It's here only as an illustration, how I actually was testing.

 arch/arm/mach-shmobile/clock-sh73a0.c |    3 +++
 1 files changed, 3 insertions(+), 0 deletions(-)

diff --git a/arch/arm/mach-shmobile/clock-sh73a0.c b/arch/arm/mach-shmobile/clock-sh73a0.c
index 3170482..f9a015d 100644
--- a/arch/arm/mach-shmobile/clock-sh73a0.c
+++ b/arch/arm/mach-shmobile/clock-sh73a0.c
@@ -809,6 +809,9 @@ void __init sh73a0_clock_init(void)
 {
 	int k, ret = 0;
 
+	if (main_clk.parent)
+		return;
+
 	/* Set SDHI clocks to a known state */
 	__raw_writel(0x108, SD0CKCR);
 	__raw_writel(0x108, SD1CKCR);
-- 
1.7.2.5

[PATCH/RFC 3/4] ARM: shmobile: sh73a0: run sh73a0_clock_init() only once

[Guennadi Liakhovetski]

sh73a0_clock_init() is called from sh73a0_earlytimer_init() and
sh73a0_add_standard_devices_dt(), take care to only run it once.

Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
---

This patch shouldn't be needed, sh73a0_clock_init() shouldn't be called 
twice. It's here only as an illustration, how I actually was testing.

 arch/arm/mach-shmobile/clock-sh73a0.c |    3 +++
 1 files changed, 3 insertions(+), 0 deletions(-)

diff --git a/arch/arm/mach-shmobile/clock-sh73a0.c b/arch/arm/mach-shmobile/clock-sh73a0.c
index 3170482..f9a015d 100644
--- a/arch/arm/mach-shmobile/clock-sh73a0.c
+++ b/arch/arm/mach-shmobile/clock-sh73a0.c
@@ -809,6 +809,9 @@ void __init sh73a0_clock_init(void)
 {
 	int k, ret = 0;
 
+	if (main_clk.parent)
+		return;
+
 	/* Set SDHI clocks to a known state */
 	__raw_writel(0x108, SD0CKCR);
 	__raw_writel(0x108, SD1CKCR);
-- 
1.7.2.5

[PATCH 4/4] ARM: shmobile: kzm9g-reference: add CPUFreq support

[Guennadi Liakhovetski]

This patch enables the use of the generic cpufreq-cpu0 driver on kzm9g.
Providing a regulator and a list of OPPs in DT, combined with a clock,
attached to the cpu0 device is everything, the cpufreq-cpu0 driver needs.

Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
---
 arch/arm/boot/dts/sh73a0-kzm9g-reference.dts |   19 ++++++++++++++++---
 arch/arm/mach-shmobile/Kconfig               |    2 ++
 2 files changed, 18 insertions(+), 3 deletions(-)

diff --git a/arch/arm/boot/dts/sh73a0-kzm9g-reference.dts b/arch/arm/boot/dts/sh73a0-kzm9g-reference.dts
index d9f6c18..ebe3f49 100644
--- a/arch/arm/boot/dts/sh73a0-kzm9g-reference.dts
+++ b/arch/arm/boot/dts/sh73a0-kzm9g-reference.dts
@@ -18,6 +18,19 @@
 	model = "KZM-A9-GT";
 	compatible = "renesas,kzm9g-reference", "renesas,sh73a0";
 
+	cpus {
+		cpu@0 {
+			cpu0-supply = <&vdd_dvfs>;
+			operating-points = <
+				/* kHz  uV */
+				1196000 1315000
+				 806000 1175000
+				 403000 1065000
+			>;
+			voltage-tolerance = <1>; /* 1% */
+		};
+	};
+
 	chosen {
 		bootargs = "console=tty0 console=ttySC4,115200 root=/dev/nfs ip=dhcp ignore_loglevel earlyprintk=sh-sci.4,115200";
 	};
@@ -118,10 +131,10 @@
 		reg = <0x40>;
 
 		regulators {
-			sd1 {
+			vdd_dvfs: sd1 {
 				regulator-name = "1.315V CPU";
-				regulator-min-microvolt = <1315000>;
-				regulator-max-microvolt = <1335000>;
+				regulator-min-microvolt = <1050000>;
+				regulator-max-microvolt = <1350000>;
 				regulator-always-on;
 				regulator-boot-on;
 			};
diff --git a/arch/arm/mach-shmobile/Kconfig b/arch/arm/mach-shmobile/Kconfig
index 927eecc..6f621ac 100644
--- a/arch/arm/mach-shmobile/Kconfig
+++ b/arch/arm/mach-shmobile/Kconfig
@@ -141,6 +141,8 @@ config MACH_KZM9G
 config MACH_KZM9G_REFERENCE
 	bool "KZM-A9-GT board - Reference Device Tree Implementation"
 	depends on ARCH_SH73A0
+	select ARCH_HAS_CPUFREQ
+	select ARCH_HAS_OPP
 	select ARCH_REQUIRE_GPIOLIB
 	select REGULATOR_FIXED_VOLTAGE if REGULATOR
 	select SND_SOC_AK4642 if SND_SIMPLE_CARD
-- 
1.7.2.5

[PATCH 4/4] ARM: shmobile: kzm9g-reference: add CPUFreq support

[Guennadi Liakhovetski]

This patch enables the use of the generic cpufreq-cpu0 driver on kzm9g.
Providing a regulator and a list of OPPs in DT, combined with a clock,
attached to the cpu0 device is everything, the cpufreq-cpu0 driver needs.

Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
---
 arch/arm/boot/dts/sh73a0-kzm9g-reference.dts |   19 ++++++++++++++++---
 arch/arm/mach-shmobile/Kconfig               |    2 ++
 2 files changed, 18 insertions(+), 3 deletions(-)

diff --git a/arch/arm/boot/dts/sh73a0-kzm9g-reference.dts b/arch/arm/boot/dts/sh73a0-kzm9g-reference.dts
index d9f6c18..ebe3f49 100644
--- a/arch/arm/boot/dts/sh73a0-kzm9g-reference.dts
+++ b/arch/arm/boot/dts/sh73a0-kzm9g-reference.dts
@@ -18,6 +18,19 @@
 	model = "KZM-A9-GT";
 	compatible = "renesas,kzm9g-reference", "renesas,sh73a0";
 
+	cpus {
+		cpu at 0 {
+			cpu0-supply = <&vdd_dvfs>;
+			operating-points = <
+				/* kHz  uV */
+				1196000 1315000
+				 806000 1175000
+				 403000 1065000
+			>;
+			voltage-tolerance = <1>; /* 1% */
+		};
+	};
+
 	chosen {
 		bootargs = "console=tty0 console=ttySC4,115200 root=/dev/nfs ip=dhcp ignore_loglevel earlyprintk=sh-sci.4,115200";
 	};
@@ -118,10 +131,10 @@
 		reg = <0x40>;
 
 		regulators {
-			sd1 {
+			vdd_dvfs: sd1 {
 				regulator-name = "1.315V CPU";
-				regulator-min-microvolt = <1315000>;
-				regulator-max-microvolt = <1335000>;
+				regulator-min-microvolt = <1050000>;
+				regulator-max-microvolt = <1350000>;
 				regulator-always-on;
 				regulator-boot-on;
 			};
diff --git a/arch/arm/mach-shmobile/Kconfig b/arch/arm/mach-shmobile/Kconfig
index 927eecc..6f621ac 100644
--- a/arch/arm/mach-shmobile/Kconfig
+++ b/arch/arm/mach-shmobile/Kconfig
@@ -141,6 +141,8 @@ config MACH_KZM9G
 config MACH_KZM9G_REFERENCE
 	bool "KZM-A9-GT board - Reference Device Tree Implementation"
 	depends on ARCH_SH73A0
+	select ARCH_HAS_CPUFREQ
+	select ARCH_HAS_OPP
 	select ARCH_REQUIRE_GPIOLIB
 	select REGULATOR_FIXED_VOLTAGE if REGULATOR
 	select SND_SOC_AK4642 if SND_SIMPLE_CARD
-- 
1.7.2.5

[PATCH 4/4] ARM: shmobile: kzm9g-reference: add CPUFreq support

[Guennadi Liakhovetski]

This patch enables the use of the generic cpufreq-cpu0 driver on kzm9g.
Providing a regulator and a list of OPPs in DT, combined with a clock,
attached to the cpu0 device is everything, the cpufreq-cpu0 driver needs.

Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
---
 arch/arm/boot/dts/sh73a0-kzm9g-reference.dts |   19 ++++++++++++++++---
 arch/arm/mach-shmobile/Kconfig               |    2 ++
 2 files changed, 18 insertions(+), 3 deletions(-)

diff --git a/arch/arm/boot/dts/sh73a0-kzm9g-reference.dts b/arch/arm/boot/dts/sh73a0-kzm9g-reference.dts
index d9f6c18..ebe3f49 100644
--- a/arch/arm/boot/dts/sh73a0-kzm9g-reference.dts
+++ b/arch/arm/boot/dts/sh73a0-kzm9g-reference.dts
@@ -18,6 +18,19 @@
 	model = "KZM-A9-GT";
 	compatible = "renesas,kzm9g-reference", "renesas,sh73a0";
 
+	cpus {
+		cpu@0 {
+			cpu0-supply = <&vdd_dvfs>;
+			operating-points = <
+				/* kHz  uV */
+				1196000 1315000
+				 806000 1175000
+				 403000 1065000
+			>;
+			voltage-tolerance = <1>; /* 1% */
+		};
+	};
+
 	chosen {
 		bootargs = "console=tty0 console=ttySC4,115200 root=/dev/nfs ip=dhcp ignore_loglevel earlyprintk=sh-sci.4,115200";
 	};
@@ -118,10 +131,10 @@
 		reg = <0x40>;
 
 		regulators {
-			sd1 {
+			vdd_dvfs: sd1 {
 				regulator-name = "1.315V CPU";
-				regulator-min-microvolt = <1315000>;
-				regulator-max-microvolt = <1335000>;
+				regulator-min-microvolt = <1050000>;
+				regulator-max-microvolt = <1350000>;
 				regulator-always-on;
 				regulator-boot-on;
 			};
diff --git a/arch/arm/mach-shmobile/Kconfig b/arch/arm/mach-shmobile/Kconfig
index 927eecc..6f621ac 100644
--- a/arch/arm/mach-shmobile/Kconfig
+++ b/arch/arm/mach-shmobile/Kconfig
@@ -141,6 +141,8 @@ config MACH_KZM9G
 config MACH_KZM9G_REFERENCE
 	bool "KZM-A9-GT board - Reference Device Tree Implementation"
 	depends on ARCH_SH73A0
+	select ARCH_HAS_CPUFREQ
+	select ARCH_HAS_OPP
 	select ARCH_REQUIRE_GPIOLIB
 	select REGULATOR_FIXED_VOLTAGE if REGULATOR
 	select SND_SOC_AK4642 if SND_SIMPLE_CARD
-- 
1.7.2.5

Re: [PATCH/RFC 0/4] CPUFreq for kzm9g

[Simon Horman]

On Fri, Feb 22, 2013 at 06:17:50PM +0100, Guennadi Liakhovetski wrote:
> The first of these patches is a bug-fix, whereas the latter three patches
> are experimental and might want to be discussed. This patch set is based
> on an earlier Simon's topic/all+next branch snapshot.

Could you let me know the commit id of the working branch?

> Attempts to use
> todays checkout failed - I couldn't even boot.

Thanks, I thought that I had tested that but clearly I was mistaken.

The problem appears to be caused by "ARM: shmobile: kzm9g: Populate
platform devices from device tree" which is present in topic/pinmux.

I will remove that branch from topic/all+next until the problem is resolved.
I have also removed the topic/eth_sh and topic/intc-of branches from
topic/all+next as they depend on topic/pinmux.

> Obviously, I'm also using 
> my AS3711 patches and a number of other (not directly related) patches, 
> that I'll be posting separately.

It is not obvious to me why you are using other patches.

> Guennadi Liakhovetski (4):
>   ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy
>   ARM: shmobile: sh73a0: add support for adjusting CPU frequency
>   ARM: shmobile: sh73a0: run sh73a0_clock_init() only once
>   ARM: shmobile: kzm9g-reference: add CPUFreq support
> 
>  arch/arm/boot/dts/sh73a0-kzm9g-reference.dts |   19 ++-
>  arch/arm/mach-shmobile/Kconfig               |    2 +
>  arch/arm/mach-shmobile/clock-sh73a0.c        |  224 +++++++++++++++++++++++++-
>  3 files changed, 236 insertions(+), 9 deletions(-)

[PATCH/RFC 0/4] CPUFreq for kzm9g

[Simon Horman]

On Fri, Feb 22, 2013 at 06:17:50PM +0100, Guennadi Liakhovetski wrote:
> The first of these patches is a bug-fix, whereas the latter three patches
> are experimental and might want to be discussed. This patch set is based
> on an earlier Simon's topic/all+next branch snapshot.

Could you let me know the commit id of the working branch?

> Attempts to use
> todays checkout failed - I couldn't even boot.

Thanks, I thought that I had tested that but clearly I was mistaken.

The problem appears to be caused by "ARM: shmobile: kzm9g: Populate
platform devices from device tree" which is present in topic/pinmux.

I will remove that branch from topic/all+next until the problem is resolved.
I have also removed the topic/eth_sh and topic/intc-of branches from
topic/all+next as they depend on topic/pinmux.

> Obviously, I'm also using 
> my AS3711 patches and a number of other (not directly related) patches, 
> that I'll be posting separately.

It is not obvious to me why you are using other patches.

> Guennadi Liakhovetski (4):
>   ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy
>   ARM: shmobile: sh73a0: add support for adjusting CPU frequency
>   ARM: shmobile: sh73a0: run sh73a0_clock_init() only once
>   ARM: shmobile: kzm9g-reference: add CPUFreq support
> 
>  arch/arm/boot/dts/sh73a0-kzm9g-reference.dts |   19 ++-
>  arch/arm/mach-shmobile/Kconfig               |    2 +
>  arch/arm/mach-shmobile/clock-sh73a0.c        |  224 +++++++++++++++++++++++++-
>  3 files changed, 236 insertions(+), 9 deletions(-)

Re: [PATCH/RFC 0/4] CPUFreq for kzm9g

[Simon Horman]

On Fri, Feb 22, 2013 at 06:17:50PM +0100, Guennadi Liakhovetski wrote:
> The first of these patches is a bug-fix, whereas the latter three patches
> are experimental and might want to be discussed. This patch set is based
> on an earlier Simon's topic/all+next branch snapshot.

Could you let me know the commit id of the working branch?

> Attempts to use
> todays checkout failed - I couldn't even boot.

Thanks, I thought that I had tested that but clearly I was mistaken.

The problem appears to be caused by "ARM: shmobile: kzm9g: Populate
platform devices from device tree" which is present in topic/pinmux.

I will remove that branch from topic/all+next until the problem is resolved.
I have also removed the topic/eth_sh and topic/intc-of branches from
topic/all+next as they depend on topic/pinmux.

> Obviously, I'm also using 
> my AS3711 patches and a number of other (not directly related) patches, 
> that I'll be posting separately.

It is not obvious to me why you are using other patches.

> Guennadi Liakhovetski (4):
>   ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy
>   ARM: shmobile: sh73a0: add support for adjusting CPU frequency
>   ARM: shmobile: sh73a0: run sh73a0_clock_init() only once
>   ARM: shmobile: kzm9g-reference: add CPUFreq support
> 
>  arch/arm/boot/dts/sh73a0-kzm9g-reference.dts |   19 ++-
>  arch/arm/mach-shmobile/Kconfig               |    2 +
>  arch/arm/mach-shmobile/clock-sh73a0.c        |  224 +++++++++++++++++++++++++-
>  3 files changed, 236 insertions(+), 9 deletions(-)

Re: [PATCH 1/4] ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy

[Simon Horman]

On Fri, Feb 22, 2013 at 06:17:51PM +0100, Guennadi Liakhovetski wrote:
> Z and ZG clocks on sh73a0 have pll0 as their parent, not pll1.

Thanks I have applied this to the soc5 branch and thus queued it up for v3.10.
Please let me know if you would prefer me to push it as a fix for v3.9.
And in that vein, if you regard it as -stable material.

> Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
> ---
>  arch/arm/mach-shmobile/clock-sh73a0.c |    4 ++--
>  1 files changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/arch/arm/mach-shmobile/clock-sh73a0.c b/arch/arm/mach-shmobile/clock-sh73a0.c
> index 5fa106b..71843dd 100644
> --- a/arch/arm/mach-shmobile/clock-sh73a0.c
> +++ b/arch/arm/mach-shmobile/clock-sh73a0.c
> @@ -265,12 +265,12 @@ enum { DIV4_I, DIV4_ZG, DIV4_M3, DIV4_B, DIV4_M1, DIV4_M2,
>  
>  static struct clk div4_clks[DIV4_NR] = {
>  	[DIV4_I] = DIV4(FRQCRA, 20, 0xdff, CLK_ENABLE_ON_INIT),
> -	[DIV4_ZG] = DIV4(FRQCRA, 16, 0xd7f, CLK_ENABLE_ON_INIT),
> +	[DIV4_ZG] = SH_CLK_DIV4(&pll0_clk, FRQCRA, 16, 0xd7f, CLK_ENABLE_ON_INIT),
>  	[DIV4_M3] = DIV4(FRQCRA, 12, 0x1dff, CLK_ENABLE_ON_INIT),
>  	[DIV4_B] = DIV4(FRQCRA, 8, 0xdff, CLK_ENABLE_ON_INIT),
>  	[DIV4_M1] = DIV4(FRQCRA, 4, 0x1dff, 0),
>  	[DIV4_M2] = DIV4(FRQCRA, 0, 0x1dff, 0),
> -	[DIV4_Z] = DIV4(FRQCRB, 24, 0x97f, 0),
> +	[DIV4_Z] = SH_CLK_DIV4(&pll0_clk, FRQCRB, 24, 0x97f, 0),
>  	[DIV4_ZTR] = DIV4(FRQCRB, 20, 0xdff, 0),
>  	[DIV4_ZT] = DIV4(FRQCRB, 16, 0xdff, 0),
>  	[DIV4_ZX] = DIV4(FRQCRB, 12, 0xdff, 0),
> -- 
> 1.7.2.5
> 

[PATCH 1/4] ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy

[Simon Horman]

On Fri, Feb 22, 2013 at 06:17:51PM +0100, Guennadi Liakhovetski wrote:
> Z and ZG clocks on sh73a0 have pll0 as their parent, not pll1.

Thanks I have applied this to the soc5 branch and thus queued it up for v3.10.
Please let me know if you would prefer me to push it as a fix for v3.9.
And in that vein, if you regard it as -stable material.

> Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
> ---
>  arch/arm/mach-shmobile/clock-sh73a0.c |    4 ++--
>  1 files changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/arch/arm/mach-shmobile/clock-sh73a0.c b/arch/arm/mach-shmobile/clock-sh73a0.c
> index 5fa106b..71843dd 100644
> --- a/arch/arm/mach-shmobile/clock-sh73a0.c
> +++ b/arch/arm/mach-shmobile/clock-sh73a0.c
> @@ -265,12 +265,12 @@ enum { DIV4_I, DIV4_ZG, DIV4_M3, DIV4_B, DIV4_M1, DIV4_M2,
>  
>  static struct clk div4_clks[DIV4_NR] = {
>  	[DIV4_I] = DIV4(FRQCRA, 20, 0xdff, CLK_ENABLE_ON_INIT),
> -	[DIV4_ZG] = DIV4(FRQCRA, 16, 0xd7f, CLK_ENABLE_ON_INIT),
> +	[DIV4_ZG] = SH_CLK_DIV4(&pll0_clk, FRQCRA, 16, 0xd7f, CLK_ENABLE_ON_INIT),
>  	[DIV4_M3] = DIV4(FRQCRA, 12, 0x1dff, CLK_ENABLE_ON_INIT),
>  	[DIV4_B] = DIV4(FRQCRA, 8, 0xdff, CLK_ENABLE_ON_INIT),
>  	[DIV4_M1] = DIV4(FRQCRA, 4, 0x1dff, 0),
>  	[DIV4_M2] = DIV4(FRQCRA, 0, 0x1dff, 0),
> -	[DIV4_Z] = DIV4(FRQCRB, 24, 0x97f, 0),
> +	[DIV4_Z] = SH_CLK_DIV4(&pll0_clk, FRQCRB, 24, 0x97f, 0),
>  	[DIV4_ZTR] = DIV4(FRQCRB, 20, 0xdff, 0),
>  	[DIV4_ZT] = DIV4(FRQCRB, 16, 0xdff, 0),
>  	[DIV4_ZX] = DIV4(FRQCRB, 12, 0xdff, 0),
> -- 
> 1.7.2.5
> 

Re: [PATCH 1/4] ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy

[Simon Horman]

On Fri, Feb 22, 2013 at 06:17:51PM +0100, Guennadi Liakhovetski wrote:
> Z and ZG clocks on sh73a0 have pll0 as their parent, not pll1.

Thanks I have applied this to the soc5 branch and thus queued it up for v3.10.
Please let me know if you would prefer me to push it as a fix for v3.9.
And in that vein, if you regard it as -stable material.

> Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
> ---
>  arch/arm/mach-shmobile/clock-sh73a0.c |    4 ++--
>  1 files changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/arch/arm/mach-shmobile/clock-sh73a0.c b/arch/arm/mach-shmobile/clock-sh73a0.c
> index 5fa106b..71843dd 100644
> --- a/arch/arm/mach-shmobile/clock-sh73a0.c
> +++ b/arch/arm/mach-shmobile/clock-sh73a0.c
> @@ -265,12 +265,12 @@ enum { DIV4_I, DIV4_ZG, DIV4_M3, DIV4_B, DIV4_M1, DIV4_M2,
>  
>  static struct clk div4_clks[DIV4_NR] = {
>  	[DIV4_I] = DIV4(FRQCRA, 20, 0xdff, CLK_ENABLE_ON_INIT),
> -	[DIV4_ZG] = DIV4(FRQCRA, 16, 0xd7f, CLK_ENABLE_ON_INIT),
> +	[DIV4_ZG] = SH_CLK_DIV4(&pll0_clk, FRQCRA, 16, 0xd7f, CLK_ENABLE_ON_INIT),
>  	[DIV4_M3] = DIV4(FRQCRA, 12, 0x1dff, CLK_ENABLE_ON_INIT),
>  	[DIV4_B] = DIV4(FRQCRA, 8, 0xdff, CLK_ENABLE_ON_INIT),
>  	[DIV4_M1] = DIV4(FRQCRA, 4, 0x1dff, 0),
>  	[DIV4_M2] = DIV4(FRQCRA, 0, 0x1dff, 0),
> -	[DIV4_Z] = DIV4(FRQCRB, 24, 0x97f, 0),
> +	[DIV4_Z] = SH_CLK_DIV4(&pll0_clk, FRQCRB, 24, 0x97f, 0),
>  	[DIV4_ZTR] = DIV4(FRQCRB, 20, 0xdff, 0),
>  	[DIV4_ZT] = DIV4(FRQCRB, 16, 0xdff, 0),
>  	[DIV4_ZX] = DIV4(FRQCRB, 12, 0xdff, 0),
> -- 
> 1.7.2.5
> 

Re: [PATCH 1/4] ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy

[Magnus Damm]

On Tue, Feb 26, 2013 at 1:29 PM, Simon Horman <horms@verge.net.au> wrote:
> On Fri, Feb 22, 2013 at 06:17:51PM +0100, Guennadi Liakhovetski wrote:
>> Z and ZG clocks on sh73a0 have pll0 as their parent, not pll1.
>
> Thanks I have applied this to the soc5 branch and thus queued it up for v3.10.
> Please let me know if you would prefer me to push it as a fix for v3.9.
> And in that vein, if you regard it as -stable material.

Thanks.

Perhaps other people disagree, but I recommend not being so trigger
happy with -stable unless it will potentially give us some upside.
AFAIK there is no in-tree consumer of this clock anyway, so involving
-stable without further testing seems like a lot of hassle with no
real benefit.

Cheers,

/ magnus

[PATCH 1/4] ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy

[Magnus Damm]

On Tue, Feb 26, 2013 at 1:29 PM, Simon Horman <horms@verge.net.au> wrote:
> On Fri, Feb 22, 2013 at 06:17:51PM +0100, Guennadi Liakhovetski wrote:
>> Z and ZG clocks on sh73a0 have pll0 as their parent, not pll1.
>
> Thanks I have applied this to the soc5 branch and thus queued it up for v3.10.
> Please let me know if you would prefer me to push it as a fix for v3.9.
> And in that vein, if you regard it as -stable material.

Thanks.

Perhaps other people disagree, but I recommend not being so trigger
happy with -stable unless it will potentially give us some upside.
AFAIK there is no in-tree consumer of this clock anyway, so involving
-stable without further testing seems like a lot of hassle with no
real benefit.

Cheers,

/ magnus

Re: [PATCH 1/4] ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy

[Magnus Damm]

On Tue, Feb 26, 2013 at 1:29 PM, Simon Horman <horms@verge.net.au> wrote:
> On Fri, Feb 22, 2013 at 06:17:51PM +0100, Guennadi Liakhovetski wrote:
>> Z and ZG clocks on sh73a0 have pll0 as their parent, not pll1.
>
> Thanks I have applied this to the soc5 branch and thus queued it up for v3.10.
> Please let me know if you would prefer me to push it as a fix for v3.9.
> And in that vein, if you regard it as -stable material.

Thanks.

Perhaps other people disagree, but I recommend not being so trigger
happy with -stable unless it will potentially give us some upside.
AFAIK there is no in-tree consumer of this clock anyway, so involving
-stable without further testing seems like a lot of hassle with no
real benefit.

Cheers,

/ magnus

Re: [PATCH 1/4] ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy

[Simon Horman]

On Tue, Feb 26, 2013 at 02:36:24PM +0900, Magnus Damm wrote:
> On Tue, Feb 26, 2013 at 1:29 PM, Simon Horman <horms@verge.net.au> wrote:
> > On Fri, Feb 22, 2013 at 06:17:51PM +0100, Guennadi Liakhovetski wrote:
> >> Z and ZG clocks on sh73a0 have pll0 as their parent, not pll1.
> >
> > Thanks I have applied this to the soc5 branch and thus queued it up for v3.10.
> > Please let me know if you would prefer me to push it as a fix for v3.9.
> > And in that vein, if you regard it as -stable material.
> 
> Thanks.
> 
> Perhaps other people disagree, but I recommend not being so trigger
> happy with -stable unless it will potentially give us some upside.
> AFAIK there is no in-tree consumer of this clock anyway, so involving
> -stable without further testing seems like a lot of hassle with no
> real benefit.

If there is no in-tree consumer then it is not v3.9 or -stable material.

Thanks!

[PATCH 1/4] ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy

[Simon Horman]

On Tue, Feb 26, 2013 at 02:36:24PM +0900, Magnus Damm wrote:
> On Tue, Feb 26, 2013 at 1:29 PM, Simon Horman <horms@verge.net.au> wrote:
> > On Fri, Feb 22, 2013 at 06:17:51PM +0100, Guennadi Liakhovetski wrote:
> >> Z and ZG clocks on sh73a0 have pll0 as their parent, not pll1.
> >
> > Thanks I have applied this to the soc5 branch and thus queued it up for v3.10.
> > Please let me know if you would prefer me to push it as a fix for v3.9.
> > And in that vein, if you regard it as -stable material.
> 
> Thanks.
> 
> Perhaps other people disagree, but I recommend not being so trigger
> happy with -stable unless it will potentially give us some upside.
> AFAIK there is no in-tree consumer of this clock anyway, so involving
> -stable without further testing seems like a lot of hassle with no
> real benefit.

If there is no in-tree consumer then it is not v3.9 or -stable material.

Thanks!

Re: [PATCH 1/4] ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy

[Simon Horman]

On Tue, Feb 26, 2013 at 02:36:24PM +0900, Magnus Damm wrote:
> On Tue, Feb 26, 2013 at 1:29 PM, Simon Horman <horms@verge.net.au> wrote:
> > On Fri, Feb 22, 2013 at 06:17:51PM +0100, Guennadi Liakhovetski wrote:
> >> Z and ZG clocks on sh73a0 have pll0 as their parent, not pll1.
> >
> > Thanks I have applied this to the soc5 branch and thus queued it up for v3.10.
> > Please let me know if you would prefer me to push it as a fix for v3.9.
> > And in that vein, if you regard it as -stable material.
> 
> Thanks.
> 
> Perhaps other people disagree, but I recommend not being so trigger
> happy with -stable unless it will potentially give us some upside.
> AFAIK there is no in-tree consumer of this clock anyway, so involving
> -stable without further testing seems like a lot of hassle with no
> real benefit.

If there is no in-tree consumer then it is not v3.9 or -stable material.

Thanks!

Re: [PATCH 1/4] ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy

[Guennadi Liakhovetski]

Hi Simon

On Tue, 26 Feb 2013, Simon Horman wrote:

> On Tue, Feb 26, 2013 at 02:36:24PM +0900, Magnus Damm wrote:
> > On Tue, Feb 26, 2013 at 1:29 PM, Simon Horman <horms@verge.net.au> wrote:
> > > On Fri, Feb 22, 2013 at 06:17:51PM +0100, Guennadi Liakhovetski wrote:
> > >> Z and ZG clocks on sh73a0 have pll0 as their parent, not pll1.
> > >
> > > Thanks I have applied this to the soc5 branch and thus queued it up for v3.10.
> > > Please let me know if you would prefer me to push it as a fix for v3.9.
> > > And in that vein, if you regard it as -stable material.
> > 
> > Thanks.
> > 
> > Perhaps other people disagree, but I recommend not being so trigger
> > happy with -stable unless it will potentially give us some upside.
> > AFAIK there is no in-tree consumer of this clock anyway, so involving
> > -stable without further testing seems like a lot of hassle with no
> > real benefit.
> 
> If there is no in-tree consumer then it is not v3.9 or -stable material.

Just to confirm, that I'm not aware of any current consumers of these 
clocks, so, no rush with stable.

Thanks
Guennadi
---
Guennadi Liakhovetski, Ph.D.
Freelance Open-Source Software Developer
http://www.open-technology.de/

[PATCH 1/4] ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy

[Guennadi Liakhovetski]

Hi Simon

On Tue, 26 Feb 2013, Simon Horman wrote:

> On Tue, Feb 26, 2013 at 02:36:24PM +0900, Magnus Damm wrote:
> > On Tue, Feb 26, 2013 at 1:29 PM, Simon Horman <horms@verge.net.au> wrote:
> > > On Fri, Feb 22, 2013 at 06:17:51PM +0100, Guennadi Liakhovetski wrote:
> > >> Z and ZG clocks on sh73a0 have pll0 as their parent, not pll1.
> > >
> > > Thanks I have applied this to the soc5 branch and thus queued it up for v3.10.
> > > Please let me know if you would prefer me to push it as a fix for v3.9.
> > > And in that vein, if you regard it as -stable material.
> > 
> > Thanks.
> > 
> > Perhaps other people disagree, but I recommend not being so trigger
> > happy with -stable unless it will potentially give us some upside.
> > AFAIK there is no in-tree consumer of this clock anyway, so involving
> > -stable without further testing seems like a lot of hassle with no
> > real benefit.
> 
> If there is no in-tree consumer then it is not v3.9 or -stable material.

Just to confirm, that I'm not aware of any current consumers of these 
clocks, so, no rush with stable.

Thanks
Guennadi
---
Guennadi Liakhovetski, Ph.D.
Freelance Open-Source Software Developer
http://www.open-technology.de/

Re: [PATCH 1/4] ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy

[Guennadi Liakhovetski]

Hi Simon

On Tue, 26 Feb 2013, Simon Horman wrote:

> On Tue, Feb 26, 2013 at 02:36:24PM +0900, Magnus Damm wrote:
> > On Tue, Feb 26, 2013 at 1:29 PM, Simon Horman <horms@verge.net.au> wrote:
> > > On Fri, Feb 22, 2013 at 06:17:51PM +0100, Guennadi Liakhovetski wrote:
> > >> Z and ZG clocks on sh73a0 have pll0 as their parent, not pll1.
> > >
> > > Thanks I have applied this to the soc5 branch and thus queued it up for v3.10.
> > > Please let me know if you would prefer me to push it as a fix for v3.9.
> > > And in that vein, if you regard it as -stable material.
> > 
> > Thanks.
> > 
> > Perhaps other people disagree, but I recommend not being so trigger
> > happy with -stable unless it will potentially give us some upside.
> > AFAIK there is no in-tree consumer of this clock anyway, so involving
> > -stable without further testing seems like a lot of hassle with no
> > real benefit.
> 
> If there is no in-tree consumer then it is not v3.9 or -stable material.

Just to confirm, that I'm not aware of any current consumers of these 
clocks, so, no rush with stable.

Thanks
Guennadi
---
Guennadi Liakhovetski, Ph.D.
Freelance Open-Source Software Developer
http://www.open-technology.de/

Re: [PATCH/RFC 0/4] CPUFreq for kzm9g

[Guennadi Liakhovetski]

(reduced cc-list)

On Mon, 25 Feb 2013, Simon Horman wrote:

> On Fri, Feb 22, 2013 at 06:17:50PM +0100, Guennadi Liakhovetski wrote:
> > The first of these patches is a bug-fix, whereas the latter three patches
> > are experimental and might want to be discussed. This patch set is based
> > on an earlier Simon's topic/all+next branch snapshot.
> 
> Could you let me know the commit id of the working branch?

It's an older snapshot (of 11.02), so, it doesn't exist on git.kernel.org 
anymore, its top commit ID is 89e9dbd.

> > Attempts to use
> > todays checkout failed - I couldn't even boot.
> 
> Thanks, I thought that I had tested that but clearly I was mistaken.
> 
> The problem appears to be caused by "ARM: shmobile: kzm9g: Populate
> platform devices from device tree" which is present in topic/pinmux.
> 
> I will remove that branch from topic/all+next until the problem is resolved.
> I have also removed the topic/eth_sh and topic/intc-of branches from
> topic/all+next as they depend on topic/pinmux.
> 
> > Obviously, I'm also using 
> > my AS3711 patches and a number of other (not directly related) patches, 
> > that I'll be posting separately.
> 
> It is not obvious to me why you are using other patches.

As I said, they are not directly related to r-/sh-mobile. Some of them are 
for cpufreq etc. Others are related, but might not be critical.

My current problem is, that the newest all+next snapshots (including the 
one from a couple of minutes ago) don't work on kzm9g either in "legacy" 
(.c-based) or in "reference" (dt-based) build. So, I am not sure against 
which tree I should be building and submitting my work. I'll try to 
identify the problem source and will let you know, unless you already know 
it. Currently kernels built from all+next typically have problems around 
bringing up the second CPU core, and even with just one core I've got a 
pinctrl-related panic later during boot.

Thanks
Guennadi

> > Guennadi Liakhovetski (4):
> >   ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy
> >   ARM: shmobile: sh73a0: add support for adjusting CPU frequency
> >   ARM: shmobile: sh73a0: run sh73a0_clock_init() only once
> >   ARM: shmobile: kzm9g-reference: add CPUFreq support
> > 
> >  arch/arm/boot/dts/sh73a0-kzm9g-reference.dts |   19 ++-
> >  arch/arm/mach-shmobile/Kconfig               |    2 +
> >  arch/arm/mach-shmobile/clock-sh73a0.c        |  224 +++++++++++++++++++++++++-
> >  3 files changed, 236 insertions(+), 9 deletions(-)
> 

---
Guennadi Liakhovetski, Ph.D.
Freelance Open-Source Software Developer
http://www.open-technology.de/

Re: [PATCH 1/4] ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy

[Simon Horman]

On Tue, Feb 26, 2013 at 07:09:00AM +0100, Guennadi Liakhovetski wrote:
> Hi Simon
> 
> On Tue, 26 Feb 2013, Simon Horman wrote:
> 
> > On Tue, Feb 26, 2013 at 02:36:24PM +0900, Magnus Damm wrote:
> > > On Tue, Feb 26, 2013 at 1:29 PM, Simon Horman <horms@verge.net.au> wrote:
> > > > On Fri, Feb 22, 2013 at 06:17:51PM +0100, Guennadi Liakhovetski wrote:
> > > >> Z and ZG clocks on sh73a0 have pll0 as their parent, not pll1.
> > > >
> > > > Thanks I have applied this to the soc5 branch and thus queued it up for v3.10.
> > > > Please let me know if you would prefer me to push it as a fix for v3.9.
> > > > And in that vein, if you regard it as -stable material.
> > > 
> > > Thanks.
> > > 
> > > Perhaps other people disagree, but I recommend not being so trigger
> > > happy with -stable unless it will potentially give us some upside.
> > > AFAIK there is no in-tree consumer of this clock anyway, so involving
> > > -stable without further testing seems like a lot of hassle with no
> > > real benefit.
> > 
> > If there is no in-tree consumer then it is not v3.9 or -stable material.
> 
> Just to confirm, that I'm not aware of any current consumers of these 
> clocks, so, no rush with stable.

Thanks. I'll leave it queued up for 3.10.

[PATCH 1/4] ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy

[Simon Horman]

On Tue, Feb 26, 2013 at 07:09:00AM +0100, Guennadi Liakhovetski wrote:
> Hi Simon
> 
> On Tue, 26 Feb 2013, Simon Horman wrote:
> 
> > On Tue, Feb 26, 2013 at 02:36:24PM +0900, Magnus Damm wrote:
> > > On Tue, Feb 26, 2013 at 1:29 PM, Simon Horman <horms@verge.net.au> wrote:
> > > > On Fri, Feb 22, 2013 at 06:17:51PM +0100, Guennadi Liakhovetski wrote:
> > > >> Z and ZG clocks on sh73a0 have pll0 as their parent, not pll1.
> > > >
> > > > Thanks I have applied this to the soc5 branch and thus queued it up for v3.10.
> > > > Please let me know if you would prefer me to push it as a fix for v3.9.
> > > > And in that vein, if you regard it as -stable material.
> > > 
> > > Thanks.
> > > 
> > > Perhaps other people disagree, but I recommend not being so trigger
> > > happy with -stable unless it will potentially give us some upside.
> > > AFAIK there is no in-tree consumer of this clock anyway, so involving
> > > -stable without further testing seems like a lot of hassle with no
> > > real benefit.
> > 
> > If there is no in-tree consumer then it is not v3.9 or -stable material.
> 
> Just to confirm, that I'm not aware of any current consumers of these 
> clocks, so, no rush with stable.

Thanks. I'll leave it queued up for 3.10.

Re: [PATCH 1/4] ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy

[Simon Horman]

On Tue, Feb 26, 2013 at 07:09:00AM +0100, Guennadi Liakhovetski wrote:
> Hi Simon
> 
> On Tue, 26 Feb 2013, Simon Horman wrote:
> 
> > On Tue, Feb 26, 2013 at 02:36:24PM +0900, Magnus Damm wrote:
> > > On Tue, Feb 26, 2013 at 1:29 PM, Simon Horman <horms@verge.net.au> wrote:
> > > > On Fri, Feb 22, 2013 at 06:17:51PM +0100, Guennadi Liakhovetski wrote:
> > > >> Z and ZG clocks on sh73a0 have pll0 as their parent, not pll1.
> > > >
> > > > Thanks I have applied this to the soc5 branch and thus queued it up for v3.10.
> > > > Please let me know if you would prefer me to push it as a fix for v3.9.
> > > > And in that vein, if you regard it as -stable material.
> > > 
> > > Thanks.
> > > 
> > > Perhaps other people disagree, but I recommend not being so trigger
> > > happy with -stable unless it will potentially give us some upside.
> > > AFAIK there is no in-tree consumer of this clock anyway, so involving
> > > -stable without further testing seems like a lot of hassle with no
> > > real benefit.
> > 
> > If there is no in-tree consumer then it is not v3.9 or -stable material.
> 
> Just to confirm, that I'm not aware of any current consumers of these 
> clocks, so, no rush with stable.

Thanks. I'll leave it queued up for 3.10.

Re: [PATCH/RFC 0/4] CPUFreq for kzm9g

[Simon Horman]

On Tue, Feb 26, 2013 at 08:19:34AM +0100, Guennadi Liakhovetski wrote:
> (reduced cc-list)
> 
> On Mon, 25 Feb 2013, Simon Horman wrote:
> 
> > On Fri, Feb 22, 2013 at 06:17:50PM +0100, Guennadi Liakhovetski wrote:
> > > The first of these patches is a bug-fix, whereas the latter three patches
> > > are experimental and might want to be discussed. This patch set is based
> > > on an earlier Simon's topic/all+next branch snapshot.
> > 
> > Could you let me know the commit id of the working branch?
> 
> It's an older snapshot (of 11.02), so, it doesn't exist on git.kernel.org 
> anymore, its top commit ID is 89e9dbd.

Thanks. I have it locally and I'll see if I can use it to
shed some light on the problem.

> > > Attempts to use
> > > todays checkout failed - I couldn't even boot.
> > 
> > Thanks, I thought that I had tested that but clearly I was mistaken.
> > 
> > The problem appears to be caused by "ARM: shmobile: kzm9g: Populate
> > platform devices from device tree" which is present in topic/pinmux.
> > 
> > I will remove that branch from topic/all+next until the problem is resolved.
> > I have also removed the topic/eth_sh and topic/intc-of branches from
> > topic/all+next as they depend on topic/pinmux.
> > 
> > > Obviously, I'm also using 
> > > my AS3711 patches and a number of other (not directly related) patches, 
> > > that I'll be posting separately.
> > 
> > It is not obvious to me why you are using other patches.
> 
> As I said, they are not directly related to r-/sh-mobile. Some of them are 
> for cpufreq etc. Others are related, but might not be critical.
> 
> My current problem is, that the newest all+next snapshots (including the 
> one from a couple of minutes ago) don't work on kzm9g either in "legacy" 
> (.c-based) or in "reference" (dt-based) build. So, I am not sure against 
> which tree I should be building and submitting my work. I'll try to 
> identify the problem source and will let you know, unless you already know 
> it. Currently kernels built from all+next typically have problems around 
> bringing up the second CPU core, and even with just one core I've got a 
> pinctrl-related panic later during boot.

Understood, I'm working on stabilising the branches.
I realise that an unstable base is difficult to work with.

> 
> Thanks
> Guennadi
> 
> > > Guennadi Liakhovetski (4):
> > >   ARM: shmobile: sh73a0: fix Z and ZG clock hierarchy
> > >   ARM: shmobile: sh73a0: add support for adjusting CPU frequency
> > >   ARM: shmobile: sh73a0: run sh73a0_clock_init() only once
> > >   ARM: shmobile: kzm9g-reference: add CPUFreq support
> > > 
> > >  arch/arm/boot/dts/sh73a0-kzm9g-reference.dts |   19 ++-
> > >  arch/arm/mach-shmobile/Kconfig               |    2 +
> > >  arch/arm/mach-shmobile/clock-sh73a0.c        |  224 +++++++++++++++++++++++++-
> > >  3 files changed, 236 insertions(+), 9 deletions(-)
> > 
> 
> ---
> Guennadi Liakhovetski, Ph.D.
> Freelance Open-Source Software Developer
> http://www.open-technology.de/
> --
> To unsubscribe from this list: send the line "unsubscribe linux-sh" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 

[PATCH 4/4] ARM: shmobile: kzm9g-reference: Remove the VCCQ MC0 function GPIO

[Laurent Pinchart]

The VCCQ MC0 power gate is now controlled by a regulator registered by
the PFC driver. Remove the corresponding function GPIO.

Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
---
 arch/arm/mach-shmobile/board-kzm9g-reference.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/arch/arm/mach-shmobile/board-kzm9g-reference.c b/arch/arm/mach-shmobile/board-kzm9g-reference.c
index aefa50d..44055fe 100644
--- a/arch/arm/mach-shmobile/board-kzm9g-reference.c
+++ b/arch/arm/mach-shmobile/board-kzm9g-reference.c
@@ -79,7 +79,6 @@ static void __init kzm_init(void)
 	sh73a0_pinmux_init();
 
 	/* enable SD */
-	gpio_request(GPIO_FN_SDHI0_VCCQ_MC0_ON,	NULL);
 	gpio_request_one(15, GPIOF_OUT_INIT_HIGH, NULL); /* power */
 
 	gpio_request_one(14, GPIOF_OUT_INIT_HIGH, NULL); /* power */
-- 
1.8.1.5