[PATCH V2 00/14] add clk_prepare/clk_unprepare to imx drivers

[PATCH V2 00/14] add clk_prepare/clk_unprepare to imx drivers

[Richard Zhao]

Changes from V1:
 - Add common helper functions clk_prepare_enable/clk_disable_unprepare
 - serial/imx: move clk_disable_unprepare before clk_put

Thanks
Richard

[PATCH V2 01/14] clk: add helper functions clk_prepare_enable and clk_disable_unprepare

[Richard Zhao]

Signed-off-by: Richard Zhao <richard.zhao-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org>
---
 include/linux/clk.h |   19 +++++++++++++++++++
 1 files changed, 19 insertions(+), 0 deletions(-)

diff --git a/include/linux/clk.h b/include/linux/clk.h
index 7213b52..63985f7 100644
--- a/include/linux/clk.h
+++ b/include/linux/clk.h
@@ -107,6 +107,25 @@ static inline void clk_unprepare(struct clk *clk)
 }
 #endif
 
+static inline int clk_prepare_enable(struct clk *clk)
+{
+	int ret;
+
+	ret = clk_prepare(clk);
+	if (!ret)
+		ret = clk_enable(clk);
+	if (ret)
+		clk_unprepare(clk);
+
+	return ret;
+}
+
+static inline void clk_disable_unprepare(struct clk *clk)
+{
+	clk_disable(clk);
+	clk_unprepare(clk);
+}
+
 /**
  * clk_get_rate - obtain the current clock rate (in Hz) for a clock source.
  *		  This is only valid once the clock source has been enabled.
-- 
1.7.5.4

[PATCH V2 02/14] ARM: mxc: time: add clk_prepare/clk_unprepare

[Richard Zhao]

Signed-off-by: Richard Zhao <richard.zhao@linaro.org>
---
 arch/arm/plat-mxc/time.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/arch/arm/plat-mxc/time.c b/arch/arm/plat-mxc/time.c
index 4b0fe28..b4e8b18 100644
--- a/arch/arm/plat-mxc/time.c
+++ b/arch/arm/plat-mxc/time.c
@@ -292,7 +292,7 @@ void __init mxc_timer_init(struct clk *timer_clk, void __iomem *base, int irq)
 {
 	uint32_t tctl_val;
 
-	clk_enable(timer_clk);
+	clk_prepare_enable(timer_clk);
 
 	timer_base = base;
 
-- 
1.7.5.4

[PATCH V2 03/14] ARM: mxc: ahci: add clk_prepare/clk_unprepare

[Richard Zhao]

Signed-off-by: Richard Zhao <richard.zhao-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org>
---
 arch/arm/plat-mxc/devices/platform-ahci-imx.c |   16 ++++++++--------
 1 files changed, 8 insertions(+), 8 deletions(-)

diff --git a/arch/arm/plat-mxc/devices/platform-ahci-imx.c b/arch/arm/plat-mxc/devices/platform-ahci-imx.c
index d8a56ae..ade4a1c 100644
--- a/arch/arm/plat-mxc/devices/platform-ahci-imx.c
+++ b/arch/arm/plat-mxc/devices/platform-ahci-imx.c
@@ -60,9 +60,9 @@ static int imx_sata_init(struct device *dev, void __iomem *addr)
 		dev_err(dev, "no sata clock.\n");
 		return PTR_ERR(sata_clk);
 	}
-	ret = clk_enable(sata_clk);
+	ret = clk_prepare_enable(sata_clk);
 	if (ret) {
-		dev_err(dev, "can't enable sata clock.\n");
+		dev_err(dev, "can't prepare/enable sata clock.\n");
 		goto put_sata_clk;
 	}
 
@@ -73,9 +73,9 @@ static int imx_sata_init(struct device *dev, void __iomem *addr)
 		ret = PTR_ERR(sata_ref_clk);
 		goto release_sata_clk;
 	}
-	ret = clk_enable(sata_ref_clk);
+	ret = clk_prepare_enable(sata_ref_clk);
 	if (ret) {
-		dev_err(dev, "can't enable sata ref clock.\n");
+		dev_err(dev, "can't prepare/enable sata ref clock.\n");
 		goto put_sata_ref_clk;
 	}
 
@@ -104,11 +104,11 @@ static int imx_sata_init(struct device *dev, void __iomem *addr)
 	return 0;
 
 release_sata_ref_clk:
-	clk_disable(sata_ref_clk);
+	clk_disable_unprepare(sata_ref_clk);
 put_sata_ref_clk:
 	clk_put(sata_ref_clk);
 release_sata_clk:
-	clk_disable(sata_clk);
+	clk_disable_unprepare(sata_clk);
 put_sata_clk:
 	clk_put(sata_clk);
 
@@ -117,10 +117,10 @@ put_sata_clk:
 
 static void imx_sata_exit(struct device *dev)
 {
-	clk_disable(sata_ref_clk);
+	clk_disable_unprepare(sata_ref_clk);
 	clk_put(sata_ref_clk);
 
-	clk_disable(sata_clk);
+	clk_disable_unprepare(sata_clk);
 	clk_put(sata_clk);
 
 }
-- 
1.7.5.4

[PATCH V2 04/14] ARM: mxc: pwm: add clk_prepare/clk_unprepare

[Richard Zhao]

Signed-off-by: Richard Zhao <richard.zhao@linaro.org>
---
 arch/arm/plat-mxc/pwm.c |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/arm/plat-mxc/pwm.c b/arch/arm/plat-mxc/pwm.c
index 42d74ea..68c7199b 100644
--- a/arch/arm/plat-mxc/pwm.c
+++ b/arch/arm/plat-mxc/pwm.c
@@ -118,7 +118,7 @@ int pwm_enable(struct pwm_device *pwm)
 	int rc = 0;
 
 	if (!pwm->clk_enabled) {
-		rc = clk_enable(pwm->clk);
+		rc = clk_prepare_enable(pwm->clk);
 		if (!rc)
 			pwm->clk_enabled = 1;
 	}
@@ -131,7 +131,7 @@ void pwm_disable(struct pwm_device *pwm)
 	writel(0, pwm->mmio_base + MX3_PWMCR);
 
 	if (pwm->clk_enabled) {
-		clk_disable(pwm->clk);
+		clk_disable_unprepare(pwm->clk);
 		pwm->clk_enabled = 0;
 	}
 }
-- 
1.7.5.4

[PATCH V2 05/14] ARM: mxc: epit: add clk_prepare/clk_unprepare

[Richard Zhao]

Signed-off-by: Richard Zhao <richard.zhao@linaro.org>
---
 arch/arm/plat-mxc/epit.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/arch/arm/plat-mxc/epit.c b/arch/arm/plat-mxc/epit.c
index d3467f8..9129c9e 100644
--- a/arch/arm/plat-mxc/epit.c
+++ b/arch/arm/plat-mxc/epit.c
@@ -203,7 +203,7 @@ static int __init epit_clockevent_init(struct clk *timer_clk)
 
 void __init epit_timer_init(struct clk *timer_clk, void __iomem *base, int irq)
 {
-	clk_enable(timer_clk);
+	clk_prepare_enable(timer_clk);
 
 	timer_base = base;
 
-- 
1.7.5.4

[PATCH V2 06/14] ARM: mxc: arch_reset: add clk_prepare/clk_unprepare

[Richard Zhao]

Signed-off-by: Richard Zhao <richard.zhao@linaro.org>
---
 arch/arm/plat-mxc/system.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/arch/arm/plat-mxc/system.c b/arch/arm/plat-mxc/system.c
index 9dad8dc..a667adc 100644
--- a/arch/arm/plat-mxc/system.c
+++ b/arch/arm/plat-mxc/system.c
@@ -54,7 +54,7 @@ void arch_reset(char mode, const char *cmd)
 
 		clk = clk_get_sys("imx2-wdt.0", NULL);
 		if (!IS_ERR(clk))
-			clk_enable(clk);
+			clk_prepare_enable(clk);
 		wcr_enable = (1 << 2);
 	}
 
-- 
1.7.5.4

[PATCH V2 07/14] ARM: mxc: audmux-v2: add clk_prepare/clk_unprepare

[Richard Zhao]

Signed-off-by: Richard Zhao <richard.zhao@linaro.org>
---
 arch/arm/plat-mxc/audmux-v2.c |    8 ++++----
 1 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/arch/arm/plat-mxc/audmux-v2.c b/arch/arm/plat-mxc/audmux-v2.c
index 8cced35..0e51fc3 100644
--- a/arch/arm/plat-mxc/audmux-v2.c
+++ b/arch/arm/plat-mxc/audmux-v2.c
@@ -73,13 +73,13 @@ static ssize_t audmux_read_file(struct file *file, char __user *user_buf,
 		return -ENOMEM;
 
 	if (audmux_clk)
-		clk_enable(audmux_clk);
+		clk_prepare_enable(audmux_clk);
 
 	ptcr = readl(audmux_base + MXC_AUDMUX_V2_PTCR(port));
 	pdcr = readl(audmux_base + MXC_AUDMUX_V2_PDCR(port));
 
 	if (audmux_clk)
-		clk_disable(audmux_clk);
+		clk_disable_unprepare(audmux_clk);
 
 	ret = snprintf(buf, PAGE_SIZE, "PDCR: %08x\nPTCR: %08x\n",
 		       pdcr, ptcr);
@@ -172,13 +172,13 @@ int mxc_audmux_v2_configure_port(unsigned int port, unsigned int ptcr,
 		return -ENOSYS;
 
 	if (audmux_clk)
-		clk_enable(audmux_clk);
+		clk_prepare_enable(audmux_clk);
 
 	writel(ptcr, audmux_base + MXC_AUDMUX_V2_PTCR(port));
 	writel(pdcr, audmux_base + MXC_AUDMUX_V2_PDCR(port));
 
 	if (audmux_clk)
-		clk_disable(audmux_clk);
+		clk_disable_unprepare(audmux_clk);
 
 	return 0;
 }
-- 
1.7.5.4

[PATCH V2 08/14] ARM: pm-imx5: add clk_prepare/clk_unprepare

[Richard Zhao]

Signed-off-by: Richard Zhao <richard.zhao@linaro.org>
---
 arch/arm/mach-mx5/pm-imx5.c |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/arm/mach-mx5/pm-imx5.c b/arch/arm/mach-mx5/pm-imx5.c
index 98052fc..2bdc85f 100644
--- a/arch/arm/mach-mx5/pm-imx5.c
+++ b/arch/arm/mach-mx5/pm-imx5.c
@@ -22,7 +22,7 @@ static struct clk *gpc_dvfs_clk;
 
 static int mx5_suspend_prepare(void)
 {
-	return clk_enable(gpc_dvfs_clk);
+	return clk_prepare_enable(gpc_dvfs_clk);
 }
 
 static int mx5_suspend_enter(suspend_state_t state)
@@ -52,7 +52,7 @@ static int mx5_suspend_enter(suspend_state_t state)
 
 static void mx5_suspend_finish(void)
 {
-	clk_disable(gpc_dvfs_clk);
+	clk_disable_unprepare(gpc_dvfs_clk);
 }
 
 static int mx5_pm_valid(suspend_state_t state)
-- 
1.7.5.4

[PATCH V2 09/14] ARM: mx31moboard: add clk_prepare/clk_unprepare

[Richard Zhao]

Signed-off-by: Richard Zhao <richard.zhao-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org>
---
 arch/arm/mach-imx/mach-mx31moboard.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/arch/arm/mach-imx/mach-mx31moboard.c b/arch/arm/mach-imx/mach-mx31moboard.c
index 07034f4..d1a9337 100644
--- a/arch/arm/mach-imx/mach-mx31moboard.c
+++ b/arch/arm/mach-imx/mach-mx31moboard.c
@@ -505,7 +505,7 @@ static void mx31moboard_poweroff(void)
 	struct clk *clk = clk_get_sys("imx2-wdt.0", NULL);
 
 	if (!IS_ERR(clk))
-		clk_enable(clk);
+		clk_prepare_enable(clk);
 
 	mxc_iomux_mode(MX31_PIN_WATCHDOG_RST__WATCHDOG_RST);
 
-- 
1.7.5.4

[PATCH V2 10/14] ARM: mxs: add clk_prepare/clk_unprepare

[Richard Zhao]

Signed-off-by: Richard Zhao <richard.zhao@linaro.org>
---
 arch/arm/mach-mxs/system.c |    2 +-
 arch/arm/mach-mxs/timer.c  |    2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/arm/mach-mxs/system.c b/arch/arm/mach-mxs/system.c
index 20ec3bd..9760a12 100644
--- a/arch/arm/mach-mxs/system.c
+++ b/arch/arm/mach-mxs/system.c
@@ -66,7 +66,7 @@ static int __init mxs_arch_reset_init(void)
 
 	clk = clk_get_sys("rtc", NULL);
 	if (!IS_ERR(clk))
-		clk_enable(clk);
+		clk_prepare_enable(clk);
 
 	return 0;
 }
diff --git a/arch/arm/mach-mxs/timer.c b/arch/arm/mach-mxs/timer.c
index cace0d2..564a632 100644
--- a/arch/arm/mach-mxs/timer.c
+++ b/arch/arm/mach-mxs/timer.c
@@ -245,7 +245,7 @@ static int __init mxs_clocksource_init(struct clk *timer_clk)
 
 void __init mxs_timer_init(struct clk *timer_clk, int irq)
 {
-	clk_enable(timer_clk);
+	clk_prepare_enable(timer_clk);
 
 	/*
 	 * Initialize timers to a known state
-- 
1.7.5.4

[PATCH V2 11/14] serial: imx: add clk_prepare/clk_unprepare

[Richard Zhao]

Signed-off-by: Richard Zhao <richard.zhao@linaro.org>
---
 drivers/tty/serial/imx.c |    7 +++----
 1 files changed, 3 insertions(+), 4 deletions(-)

diff --git a/drivers/tty/serial/imx.c b/drivers/tty/serial/imx.c
index 163fc90..8f3709f 100644
--- a/drivers/tty/serial/imx.c
+++ b/drivers/tty/serial/imx.c
@@ -1390,7 +1390,7 @@ static int serial_imx_probe(struct platform_device *pdev)
 		ret = PTR_ERR(sport->clk);
 		goto unmap;
 	}
-	clk_enable(sport->clk);
+	clk_prepare_enable(sport->clk);
 
 	sport->port.uartclk = clk_get_rate(sport->clk);
 
@@ -1413,8 +1413,8 @@ deinit:
 	if (pdata && pdata->exit)
 		pdata->exit(pdev);
 clkput:
+	clk_disable_unprepare(sport->clk);
 	clk_put(sport->clk);
-	clk_disable(sport->clk);
 unmap:
 	iounmap(sport->port.membase);
 free:
@@ -1434,11 +1434,10 @@ static int serial_imx_remove(struct platform_device *pdev)
 
 	if (sport) {
 		uart_remove_one_port(&imx_reg, &sport->port);
+		clk_disable_unprepare(sport->clk);
 		clk_put(sport->clk);
 	}
 
-	clk_disable(sport->clk);
-
 	if (pdata && pdata->exit)
 		pdata->exit(pdev);
 
-- 
1.7.5.4

[PATCH V2 12/14] net: fec: add clk_prepare/clk_unprepare

[Richard Zhao]

Signed-off-by: Richard Zhao <richard.zhao@linaro.org>
---
 drivers/net/ethernet/freescale/fec.c |   10 +++++-----
 1 files changed, 5 insertions(+), 5 deletions(-)

diff --git a/drivers/net/ethernet/freescale/fec.c b/drivers/net/ethernet/freescale/fec.c
index 1124ce0..e96fa44 100644
--- a/drivers/net/ethernet/freescale/fec.c
+++ b/drivers/net/ethernet/freescale/fec.c
@@ -1588,7 +1588,7 @@ fec_probe(struct platform_device *pdev)
 		ret = PTR_ERR(fep->clk);
 		goto failed_clk;
 	}
-	clk_enable(fep->clk);
+	clk_prepare_enable(fep->clk);
 
 	ret = fec_enet_init(ndev);
 	if (ret)
@@ -1611,7 +1611,7 @@ failed_register:
 	fec_enet_mii_remove(fep);
 failed_mii_init:
 failed_init:
-	clk_disable(fep->clk);
+	clk_disable_unprepare(fep->clk);
 	clk_put(fep->clk);
 failed_clk:
 	for (i = 0; i < FEC_IRQ_NUM; i++) {
@@ -1638,7 +1638,7 @@ fec_drv_remove(struct platform_device *pdev)
 
 	fec_stop(ndev);
 	fec_enet_mii_remove(fep);
-	clk_disable(fep->clk);
+	clk_disable_unprepare(fep->clk);
 	clk_put(fep->clk);
 	iounmap(fep->hwp);
 	unregister_netdev(ndev);
@@ -1664,7 +1664,7 @@ fec_suspend(struct device *dev)
 		fec_stop(ndev);
 		netif_device_detach(ndev);
 	}
-	clk_disable(fep->clk);
+	clk_disable_unprepare(fep->clk);
 
 	return 0;
 }
@@ -1675,7 +1675,7 @@ fec_resume(struct device *dev)
 	struct net_device *ndev = dev_get_drvdata(dev);
 	struct fec_enet_private *fep = netdev_priv(ndev);
 
-	clk_enable(fep->clk);
+	clk_prepare_enable(fep->clk);
 	if (netif_running(ndev)) {
 		fec_restart(ndev, fep->full_duplex);
 		netif_device_attach(ndev);
-- 
1.7.5.4

[PATCH V2 13/14] i2c: imx: add clk_prepare/clk_unprepare

[Richard Zhao]

Signed-off-by: Richard Zhao <richard.zhao@linaro.org>
---
 drivers/i2c/busses/i2c-imx.c |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/i2c/busses/i2c-imx.c b/drivers/i2c/busses/i2c-imx.c
index 58832e5..8d1ab6f 100644
--- a/drivers/i2c/busses/i2c-imx.c
+++ b/drivers/i2c/busses/i2c-imx.c
@@ -196,7 +196,7 @@ static int i2c_imx_start(struct imx_i2c_struct *i2c_imx)
 
 	dev_dbg(&i2c_imx->adapter.dev, "<%s>\n", __func__);
 
-	clk_enable(i2c_imx->clk);
+	clk_prepare_enable(i2c_imx->clk);
 	writeb(i2c_imx->ifdr, i2c_imx->base + IMX_I2C_IFDR);
 	/* Enable I2C controller */
 	writeb(0, i2c_imx->base + IMX_I2C_I2SR);
@@ -245,7 +245,7 @@ static void i2c_imx_stop(struct imx_i2c_struct *i2c_imx)
 
 	/* Disable I2C controller */
 	writeb(0, i2c_imx->base + IMX_I2C_I2CR);
-	clk_disable(i2c_imx->clk);
+	clk_disable_unprepare(i2c_imx->clk);
 }
 
 static void __init i2c_imx_set_clk(struct imx_i2c_struct *i2c_imx,
-- 
1.7.5.4

[PATCH V2 14/14] mmc: sdhci-esdhc-imx: add clk_prepare/clk_unprepare

[Richard Zhao]

Signed-off-by: Richard Zhao <richard.zhao@linaro.org>
---
 drivers/mmc/host/sdhci-esdhc-imx.c |    6 +++---
 1 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/mmc/host/sdhci-esdhc-imx.c b/drivers/mmc/host/sdhci-esdhc-imx.c
index ae57769..f1ae37e 100644
--- a/drivers/mmc/host/sdhci-esdhc-imx.c
+++ b/drivers/mmc/host/sdhci-esdhc-imx.c
@@ -462,7 +462,7 @@ static int __devinit sdhci_esdhc_imx_probe(struct platform_device *pdev)
 		err = PTR_ERR(clk);
 		goto err_clk_get;
 	}
-	clk_enable(clk);
+	clk_prepare_enable(clk);
 	pltfm_host->clk = clk;
 
 	if (!is_imx25_esdhc(imx_data))
@@ -550,7 +550,7 @@ no_card_detect_irq:
 		gpio_free(boarddata->wp_gpio);
 no_card_detect_pin:
 no_board_data:
-	clk_disable(pltfm_host->clk);
+	clk_disable_unprepare(pltfm_host->clk);
 	clk_put(pltfm_host->clk);
 err_clk_get:
 	kfree(imx_data);
@@ -577,7 +577,7 @@ static int __devexit sdhci_esdhc_imx_remove(struct platform_device *pdev)
 		gpio_free(boarddata->cd_gpio);
 	}
 
-	clk_disable(pltfm_host->clk);
+	clk_disable_unprepare(pltfm_host->clk);
 	clk_put(pltfm_host->clk);
 	kfree(imx_data);
 
-- 
1.7.5.4

Re: [PATCH V2 01/14] clk: add helper functions clk_prepare_enable and clk_disable_unprepare

[Baruch Siach]

Hi Richard,

On Fri, Nov 11, 2011 at 09:10:16AM +0800, Richard Zhao wrote:
> Signed-off-by: Richard Zhao <richard.zhao@linaro.org>
> ---
>  include/linux/clk.h |   19 +++++++++++++++++++
>  1 files changed, 19 insertions(+), 0 deletions(-)
> 
> diff --git a/include/linux/clk.h b/include/linux/clk.h
> index 7213b52..63985f7 100644
> --- a/include/linux/clk.h
> +++ b/include/linux/clk.h
> @@ -107,6 +107,25 @@ static inline void clk_unprepare(struct clk *clk)
>  }
>  #endif
>  
> +static inline int clk_prepare_enable(struct clk *clk)
> +{
> +	int ret;
> +
> +	ret = clk_prepare(clk);
> +	if (!ret)
> +		ret = clk_enable(clk);
> +	if (ret)

Using 'else' here makes the code more readable IMO.

baruch

> +		clk_unprepare(clk);
> +
> +	return ret;
> +}
> +
> +static inline void clk_disable_unprepare(struct clk *clk)
> +{
> +	clk_disable(clk);
> +	clk_unprepare(clk);
> +}
> +
>  /**
>   * clk_get_rate - obtain the current clock rate (in Hz) for a clock source.
>   *		  This is only valid once the clock source has been enabled.
> -- 
> 1.7.5.4

-- 
                                                     ~. .~   Tk Open Systems
=}------------------------------------------------ooO--U--Ooo------------{=
   - baruch@tkos.co.il - tel: +972.2.679.5364, http://www.tkos.co.il -

Re: [PATCH V2 01/14] clk: add helper functions clk_prepare_enable and clk_disable_unprepare

[Sascha Hauer]

On Fri, Nov 11, 2011 at 09:10:16AM +0800, Richard Zhao wrote:
> Signed-off-by: Richard Zhao <richard.zhao@linaro.org>
> ---
>  include/linux/clk.h |   19 +++++++++++++++++++
>  1 files changed, 19 insertions(+), 0 deletions(-)
> 
> diff --git a/include/linux/clk.h b/include/linux/clk.h
> index 7213b52..63985f7 100644
> --- a/include/linux/clk.h
> +++ b/include/linux/clk.h
> @@ -107,6 +107,25 @@ static inline void clk_unprepare(struct clk *clk)
>  }
>  #endif
>  
> +static inline int clk_prepare_enable(struct clk *clk)
> +{
> +	int ret;
> +
> +	ret = clk_prepare(clk);
> +	if (!ret)
> +		ret = clk_enable(clk);
> +	if (ret)
> +		clk_unprepare(clk);

As Baruch already pointed out this is hard to read. Also it contains a
bug. When clk_prepare fails clk_unprepare is called afterwards.

Hint: Kernel guys are not afraid of having multiple return statements
in a function, instead they like it when something is handled early
without having to read the rest of the function.

Sascha

-- 
Pengutronix e.K.                           |                             |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |
Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |

Re: [PATCH V2 01/14] clk: add helper functions clk_prepare_enable and clk_disable_unprepare

[Richard Zhao]

On Fri, Nov 11, 2011 at 09:49:50AM +0100, Sascha Hauer wrote:
> On Fri, Nov 11, 2011 at 09:10:16AM +0800, Richard Zhao wrote:
> > Signed-off-by: Richard Zhao <richard.zhao@linaro.org>
> > ---
> >  include/linux/clk.h |   19 +++++++++++++++++++
> >  1 files changed, 19 insertions(+), 0 deletions(-)
> > 
> > diff --git a/include/linux/clk.h b/include/linux/clk.h
> > index 7213b52..63985f7 100644
> > --- a/include/linux/clk.h
> > +++ b/include/linux/clk.h
> > @@ -107,6 +107,25 @@ static inline void clk_unprepare(struct clk *clk)
> >  }
> >  #endif
> >  
> > +static inline int clk_prepare_enable(struct clk *clk)
> > +{
> > +	int ret;
> > +
> > +	ret = clk_prepare(clk);
> > +	if (!ret)
> > +		ret = clk_enable(clk);
> > +	if (ret)
> > +		clk_unprepare(clk);
> 
> As Baruch already pointed out this is hard to read. Also it contains a
> bug. When clk_prepare fails clk_unprepare is called afterwards.
> 
> Hint: Kernel guys are not afraid of having multiple return statements
> in a function, instead they like it when something is handled early
> without having to read the rest of the function.
{
	int ret;

	ret = clk_prepare(clk);
	if (ret)
		return ret;
	ret = clk_enable(clk);
	if (ret)
		clk_unprepare(clk);
	return ret;

Like above code? Thanks for review.

Richard
> 
> Sascha
> 
> -- 
> Pengutronix e.K.                           |                             |
> Industrial Linux Solutions                 | http://www.pengutronix.de/  |
> Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0    |
> Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |
> 
> _______________________________________________
> linux-arm-kernel mailing list
> linux-arm-kernel@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
> 

Re: [PATCH V2 01/14] clk: add helper functions clk_prepare_enable and clk_disable_unprepare

[Sascha Hauer]

On Fri, Nov 11, 2011 at 05:05:47PM +0800, Richard Zhao wrote:
> On Fri, Nov 11, 2011 at 09:49:50AM +0100, Sascha Hauer wrote:
> > On Fri, Nov 11, 2011 at 09:10:16AM +0800, Richard Zhao wrote:
> > > Signed-off-by: Richard Zhao <richard.zhao@linaro.org>
> > > ---
> > >  include/linux/clk.h |   19 +++++++++++++++++++
> > >  1 files changed, 19 insertions(+), 0 deletions(-)
> > > 
> > > diff --git a/include/linux/clk.h b/include/linux/clk.h
> > > index 7213b52..63985f7 100644
> > > --- a/include/linux/clk.h
> > > +++ b/include/linux/clk.h
> > > @@ -107,6 +107,25 @@ static inline void clk_unprepare(struct clk *clk)
> > >  }
> > >  #endif
> > >  
> > > +static inline int clk_prepare_enable(struct clk *clk)
> > > +{
> > > +	int ret;
> > > +
> > > +	ret = clk_prepare(clk);
> > > +	if (!ret)
> > > +		ret = clk_enable(clk);
> > > +	if (ret)
> > > +		clk_unprepare(clk);
> > 
> > As Baruch already pointed out this is hard to read. Also it contains a
> > bug. When clk_prepare fails clk_unprepare is called afterwards.
> > 
> > Hint: Kernel guys are not afraid of having multiple return statements
> > in a function, instead they like it when something is handled early
> > without having to read the rest of the function.
> {
> 	int ret;
> 
> 	ret = clk_prepare(clk);
> 	if (ret)
> 		return ret;
> 	ret = clk_enable(clk);
> 	if (ret)
> 		clk_unprepare(clk);
> 	return ret;

Yes, looks good.

Sascha

-- 
Pengutronix e.K.                           |                             |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |
Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |

Re: [PATCH V2 01/14] clk: add helper functions clk_prepare_enable and clk_disable_unprepare

[Russell King - ARM Linux]

On Fri, Nov 11, 2011 at 10:15:56AM +0100, Sascha Hauer wrote:
> On Fri, Nov 11, 2011 at 05:05:47PM +0800, Richard Zhao wrote:
> > {
> > 	int ret;
> > 
> > 	ret = clk_prepare(clk);
> > 	if (ret)
> > 		return ret;
> > 	ret = clk_enable(clk);
> > 	if (ret)
> > 		clk_unprepare(clk);
> > 	return ret;
> 
> Yes, looks good.

While this looks like a nice easy solution for converting existing
drivers, I'd suggest thinking about this a little more...

I would suggest some thought is given to the placement of clk_enable()
and clk_disable() when adding clk_prepare(), especially if your existing
clk_enable() function can only be called from non-atomic contexts.

Obviously, the transition path needs to be along these lines:

1. add clk_prepare() to drivers
2. implement clk_prepare() and make clk_enable() callable from non-atomic
   contexts
3. move clk_enable() in drivers to places it can be called from non-atomic
   contexts to achieve greater power savings (maybe via the runtime pm)

and where a driver is shared between different sub-architectures which
have non-atomic clk_enable()s, (3) can only happen when all those sub-
architectures have been updated to step (2).

Re: [PATCH V2 01/14] clk: add helper functions clk_prepare_enable and clk_disable_unprepare

[Sascha Hauer]

On Fri, Nov 11, 2011 at 10:27:27AM +0000, Russell King - ARM Linux wrote:
> On Fri, Nov 11, 2011 at 10:15:56AM +0100, Sascha Hauer wrote:
> > On Fri, Nov 11, 2011 at 05:05:47PM +0800, Richard Zhao wrote:
> > > {
> > > 	int ret;
> > > 
> > > 	ret = clk_prepare(clk);
> > > 	if (ret)
> > > 		return ret;
> > > 	ret = clk_enable(clk);
> > > 	if (ret)
> > > 		clk_unprepare(clk);
> > > 	return ret;
> > 
> > Yes, looks good.
> 
> While this looks like a nice easy solution for converting existing
> drivers, I'd suggest thinking about this a little more...
> 
> I would suggest some thought is given to the placement of clk_enable()
> and clk_disable() when adding clk_prepare(), especially if your existing
> clk_enable() function can only be called from non-atomic contexts.
> 
> Obviously, the transition path needs to be along these lines:
> 
> 1. add clk_prepare() to drivers
> 2. implement clk_prepare() and make clk_enable() callable from non-atomic
>    contexts
> 3. move clk_enable() in drivers to places it can be called from non-atomic
>    contexts to achieve greater power savings (maybe via the runtime pm)
> 
> and where a driver is shared between different sub-architectures which
> have non-atomic clk_enable()s, (3) can only happen when all those sub-
> architectures have been updated to step (2).

The drivers changed here all do clk_prepare/enable in their probe
function. I agree that this clk_prepare_enable patch gives kind of
wrong motivation to just use this function and to forget about
potential power savings with proper integration of clk_prepare/enable.
I think though that it will take a long time until all drivers really
do this no matter if we have such a helper or not. I think that in the
meantime it's better to have a little helper than to clobber the probe
code with additional error handling.

Sascha

-- 
Pengutronix e.K.                           |                             |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |
Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |

Re: [PATCH V2 01/14] clk: add helper functions clk_prepare_enable and clk_disable_unprepare

[Richard Zhao]

On Fri, Nov 11, 2011 at 12:56:44PM +0100, Sascha Hauer wrote:
> On Fri, Nov 11, 2011 at 10:27:27AM +0000, Russell King - ARM Linux wrote:
> > On Fri, Nov 11, 2011 at 10:15:56AM +0100, Sascha Hauer wrote:
> > > On Fri, Nov 11, 2011 at 05:05:47PM +0800, Richard Zhao wrote:
> > > > {
> > > > 	int ret;
> > > > 
> > > > 	ret = clk_prepare(clk);
> > > > 	if (ret)
> > > > 		return ret;
> > > > 	ret = clk_enable(clk);
> > > > 	if (ret)
> > > > 		clk_unprepare(clk);
> > > > 	return ret;
> > > 
> > > Yes, looks good.
> > 
> > While this looks like a nice easy solution for converting existing
> > drivers, I'd suggest thinking about this a little more...
> > 
> > I would suggest some thought is given to the placement of clk_enable()
> > and clk_disable() when adding clk_prepare(), especially if your existing
> > clk_enable() function can only be called from non-atomic contexts.
> > 
> > Obviously, the transition path needs to be along these lines:
> > 
> > 1. add clk_prepare() to drivers
> > 2. implement clk_prepare() and make clk_enable() callable from non-atomic
> >    contexts
> > 3. move clk_enable() in drivers to places it can be called from non-atomic
> >    contexts to achieve greater power savings (maybe via the runtime pm)
Hi Russell,

There are use cases calling clk_enable in atomic context, but there are even more
cases caling in non-atomic context. The patch meant to help the latter cases.
Tuning of driver power savings can be left to other contributers. Is it ok
if I add below comments:
/* clk_prepare_enable helps cases using clk_enable in non-atomic context. */
static inline int clk_prepare_enable(struct clk *clk)
...
/* clk_disable_unprepare helps cases using clk_disable in non-atomic context. */
static inline void clk_disable_unprepare(struct clk *clk)
...

Thanks
Richard
> > 
> > and where a driver is shared between different sub-architectures which
> > have non-atomic clk_enable()s, (3) can only happen when all those sub-
> > architectures have been updated to step (2).
> 
> The drivers changed here all do clk_prepare/enable in their probe
> function. I agree that this clk_prepare_enable patch gives kind of
> wrong motivation to just use this function and to forget about
> potential power savings with proper integration of clk_prepare/enable.
> I think though that it will take a long time until all drivers really
> do this no matter if we have such a helper or not. I think that in the
> meantime it's better to have a little helper than to clobber the probe
> code with additional error handling.
> 
> Sascha
> 
> -- 
> Pengutronix e.K.                           |                             |
> Industrial Linux Solutions                 | http://www.pengutronix.de/  |
> Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0    |
> Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |
> 
> _______________________________________________
> linux-arm-kernel mailing list
> linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r@public.gmane.org
> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
>