diff mbox series

[resend] mtd: maps: physmap: Add minimal Runtime PM support

Message ID 20200115131323.6883-1-geert+renesas@glider.be (mailing list archive)
State Mainlined
Commit 0bc448b49e8a017e16edf843baf5b4221e191b1f
Delegated to: Geert Uytterhoeven
Headers show
Series [resend] mtd: maps: physmap: Add minimal Runtime PM support | expand

Commit Message

Geert Uytterhoeven Jan. 15, 2020, 1:13 p.m. UTC
Add minimal runtime PM support (enable on probe, disable on remove), to
ensure proper operation with a parent device that uses runtime PM.

This is needed on systems where the FLASH is connected to a bus
controller that is contained in a PM domain and/or has a gateable
functional clock.  In such cases, before accessing any device connected
to the external bus, the PM domain must be powered up, and/or the
functional clock must be enabled, which is typically handled through
runtime PM by the bus controller driver.

An example of this is the Renesas APE6-EVM development board, which has
an Ethernet controller and a CFI FLASH connected to the Bus State
Controller (BSC) of an R-Mobile APE6 SoC.
As long as the Ethernet driver, which had Runtime PM support since
commit 3a611e26e958b037 ("net/smsc911x: Add minimal runtime PM
support"), keeps the BSC powered, accessing the FLASH works.
When the ethernet node in r8a73a4-ape6evm.dts is disabled, the BSC is
never powered up, and the kernel crashes when trying to access the
FLASH:

    Unhandled fault: imprecise external abort (0x1406) at 0x00000000
    pgd = (ptrval)
    [00000000] *pgd=7fef2835
    Internal error: : 1406 [#1] SMP ARM
    CPU: 0 PID: 122 Comm: hd Tainted: G        W         5.5.0-rc1-ape6evm-00814-g38ca966db25b9dbd-dirty #136
    Hardware name: Generic R8A73A4 (Flattened Device Tree)
    PC is at chip_ready+0x12c/0x380
    LR is at chip_ready+0x10c/0x380

Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
---
Probably the device should be powered down after probing, and powered
up/down on-demand in the various {get,put}_chip() functions.  However,
that is an optimization which touches more intimidate details of the
internal MTD API, and can be done later.
---
 drivers/mtd/maps/physmap-core.c | 20 +++++++++++++++-----
 1 file changed, 15 insertions(+), 5 deletions(-)

Comments

Niklas Söderlund Jan. 15, 2020, 3:42 p.m. UTC | #1
Hi Geert,

Thanks for your work.

On 2020-01-15 14:13:23 +0100, Geert Uytterhoeven wrote:
> Add minimal runtime PM support (enable on probe, disable on remove), to
> ensure proper operation with a parent device that uses runtime PM.
> 
> This is needed on systems where the FLASH is connected to a bus
> controller that is contained in a PM domain and/or has a gateable
> functional clock.  In such cases, before accessing any device connected
> to the external bus, the PM domain must be powered up, and/or the
> functional clock must be enabled, which is typically handled through
> runtime PM by the bus controller driver.
> 
> An example of this is the Renesas APE6-EVM development board, which has
> an Ethernet controller and a CFI FLASH connected to the Bus State
> Controller (BSC) of an R-Mobile APE6 SoC.
> As long as the Ethernet driver, which had Runtime PM support since
> commit 3a611e26e958b037 ("net/smsc911x: Add minimal runtime PM
> support"), keeps the BSC powered, accessing the FLASH works.
> When the ethernet node in r8a73a4-ape6evm.dts is disabled, the BSC is
> never powered up, and the kernel crashes when trying to access the
> FLASH:
> 
>     Unhandled fault: imprecise external abort (0x1406) at 0x00000000
>     pgd = (ptrval)
>     [00000000] *pgd=7fef2835
>     Internal error: : 1406 [#1] SMP ARM
>     CPU: 0 PID: 122 Comm: hd Tainted: G        W         5.5.0-rc1-ape6evm-00814-g38ca966db25b9dbd-dirty #136
>     Hardware name: Generic R8A73A4 (Flattened Device Tree)
>     PC is at chip_ready+0x12c/0x380
>     LR is at chip_ready+0x10c/0x380
> 
> Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>

Reviewed-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>

> ---
> Probably the device should be powered down after probing, and powered
> up/down on-demand in the various {get,put}_chip() functions.  However,
> that is an optimization which touches more intimidate details of the
> internal MTD API, and can be done later.
> ---
>  drivers/mtd/maps/physmap-core.c | 20 +++++++++++++++-----
>  1 file changed, 15 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/mtd/maps/physmap-core.c b/drivers/mtd/maps/physmap-core.c
> index a9f7964e2edb6668..8f7f966fa9a7ee8a 100644
> --- a/drivers/mtd/maps/physmap-core.c
> +++ b/drivers/mtd/maps/physmap-core.c
> @@ -38,6 +38,7 @@
>  #include <linux/mtd/cfi_endian.h>
>  #include <linux/io.h>
>  #include <linux/of_device.h>
> +#include <linux/pm_runtime.h>
>  #include <linux/gpio/consumer.h>
>  
>  #include "physmap-gemini.h"
> @@ -64,16 +65,16 @@ static int physmap_flash_remove(struct platform_device *dev)
>  {
>  	struct physmap_flash_info *info;
>  	struct physmap_flash_data *physmap_data;
> -	int i, err;
> +	int i, err = 0;
>  
>  	info = platform_get_drvdata(dev);
>  	if (!info)
> -		return 0;
> +		goto out;
>  
>  	if (info->cmtd) {
>  		err = mtd_device_unregister(info->cmtd);
>  		if (err)
> -			return err;
> +			goto out;
>  
>  		if (info->cmtd != info->mtds[0])
>  			mtd_concat_destroy(info->cmtd);
> @@ -88,7 +89,10 @@ static int physmap_flash_remove(struct platform_device *dev)
>  	if (physmap_data && physmap_data->exit)
>  		physmap_data->exit(dev);
>  
> -	return 0;
> +out:
> +	pm_runtime_put(&dev->dev);
> +	pm_runtime_disable(&dev->dev);
> +	return err;
>  }
>  
>  static void physmap_set_vpp(struct map_info *map, int state)
> @@ -484,13 +488,19 @@ static int physmap_flash_probe(struct platform_device *dev)
>  		return -EINVAL;
>  	}
>  
> +	pm_runtime_enable(&dev->dev);
> +	pm_runtime_get_sync(&dev->dev);
> +
>  	if (dev->dev.of_node)
>  		err = physmap_flash_of_init(dev);
>  	else
>  		err = physmap_flash_pdata_init(dev);
>  
> -	if (err)
> +	if (err) {
> +		pm_runtime_put(&dev->dev);
> +		pm_runtime_disable(&dev->dev);
>  		return err;
> +	}
>  
>  	for (i = 0; i < info->nmaps; i++) {
>  		struct resource *res;
> -- 
> 2.17.1
>
Kevin Hilman Jan. 15, 2020, 11:48 p.m. UTC | #2
Geert Uytterhoeven <geert+renesas@glider.be> writes:

> Add minimal runtime PM support (enable on probe, disable on remove), to
> ensure proper operation with a parent device that uses runtime PM.
>
> This is needed on systems where the FLASH is connected to a bus
> controller that is contained in a PM domain and/or has a gateable
> functional clock.  In such cases, before accessing any device connected
> to the external bus, the PM domain must be powered up, and/or the
> functional clock must be enabled, which is typically handled through
> runtime PM by the bus controller driver.
>
> An example of this is the Renesas APE6-EVM development board, which has
> an Ethernet controller and a CFI FLASH connected to the Bus State
> Controller (BSC) of an R-Mobile APE6 SoC.
> As long as the Ethernet driver, which had Runtime PM support since
> commit 3a611e26e958b037 ("net/smsc911x: Add minimal runtime PM
> support"), keeps the BSC powered, accessing the FLASH works.
> When the ethernet node in r8a73a4-ape6evm.dts is disabled, the BSC is
> never powered up, and the kernel crashes when trying to access the
> FLASH:
>
>     Unhandled fault: imprecise external abort (0x1406) at 0x00000000
>     pgd = (ptrval)
>     [00000000] *pgd=7fef2835
>     Internal error: : 1406 [#1] SMP ARM
>     CPU: 0 PID: 122 Comm: hd Tainted: G        W         5.5.0-rc1-ape6evm-00814-g38ca966db25b9dbd-dirty #136
>     Hardware name: Generic R8A73A4 (Flattened Device Tree)
>     PC is at chip_ready+0x12c/0x380
>     LR is at chip_ready+0x10c/0x380
>
> Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>

Reviewed-by: Kevin Hilman <khilman@baylibre.com>

> ---
> Probably the device should be powered down after probing, and powered
> up/down on-demand in the various {get,put}_chip() functions.  However,
> that is an optimization which touches more intimidate details of the
> internal MTD API, and can be done later.

Agreed.  Without that, the BSC (in your example) will never get powered
off as long as there's a flash device pobed.

Kevin

> ---
>  drivers/mtd/maps/physmap-core.c | 20 +++++++++++++++-----
>  1 file changed, 15 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/mtd/maps/physmap-core.c b/drivers/mtd/maps/physmap-core.c
> index a9f7964e2edb6668..8f7f966fa9a7ee8a 100644
> --- a/drivers/mtd/maps/physmap-core.c
> +++ b/drivers/mtd/maps/physmap-core.c
> @@ -38,6 +38,7 @@
>  #include <linux/mtd/cfi_endian.h>
>  #include <linux/io.h>
>  #include <linux/of_device.h>
> +#include <linux/pm_runtime.h>
>  #include <linux/gpio/consumer.h>
>  
>  #include "physmap-gemini.h"
> @@ -64,16 +65,16 @@ static int physmap_flash_remove(struct platform_device *dev)
>  {
>  	struct physmap_flash_info *info;
>  	struct physmap_flash_data *physmap_data;
> -	int i, err;
> +	int i, err = 0;
>  
>  	info = platform_get_drvdata(dev);
>  	if (!info)
> -		return 0;
> +		goto out;
>  
>  	if (info->cmtd) {
>  		err = mtd_device_unregister(info->cmtd);
>  		if (err)
> -			return err;
> +			goto out;
>  
>  		if (info->cmtd != info->mtds[0])
>  			mtd_concat_destroy(info->cmtd);
> @@ -88,7 +89,10 @@ static int physmap_flash_remove(struct platform_device *dev)
>  	if (physmap_data && physmap_data->exit)
>  		physmap_data->exit(dev);
>  
> -	return 0;
> +out:
> +	pm_runtime_put(&dev->dev);
> +	pm_runtime_disable(&dev->dev);
> +	return err;
>  }
>  
>  static void physmap_set_vpp(struct map_info *map, int state)
> @@ -484,13 +488,19 @@ static int physmap_flash_probe(struct platform_device *dev)
>  		return -EINVAL;
>  	}
>  
> +	pm_runtime_enable(&dev->dev);
> +	pm_runtime_get_sync(&dev->dev);
> +
>  	if (dev->dev.of_node)
>  		err = physmap_flash_of_init(dev);
>  	else
>  		err = physmap_flash_pdata_init(dev);
>  
> -	if (err)
> +	if (err) {
> +		pm_runtime_put(&dev->dev);
> +		pm_runtime_disable(&dev->dev);
>  		return err;
> +	}
>  
>  	for (i = 0; i < info->nmaps; i++) {
>  		struct resource *res;
> -- 
> 2.17.1
diff mbox series

Patch

diff --git a/drivers/mtd/maps/physmap-core.c b/drivers/mtd/maps/physmap-core.c
index a9f7964e2edb6668..8f7f966fa9a7ee8a 100644
--- a/drivers/mtd/maps/physmap-core.c
+++ b/drivers/mtd/maps/physmap-core.c
@@ -38,6 +38,7 @@ 
 #include <linux/mtd/cfi_endian.h>
 #include <linux/io.h>
 #include <linux/of_device.h>
+#include <linux/pm_runtime.h>
 #include <linux/gpio/consumer.h>
 
 #include "physmap-gemini.h"
@@ -64,16 +65,16 @@  static int physmap_flash_remove(struct platform_device *dev)
 {
 	struct physmap_flash_info *info;
 	struct physmap_flash_data *physmap_data;
-	int i, err;
+	int i, err = 0;
 
 	info = platform_get_drvdata(dev);
 	if (!info)
-		return 0;
+		goto out;
 
 	if (info->cmtd) {
 		err = mtd_device_unregister(info->cmtd);
 		if (err)
-			return err;
+			goto out;
 
 		if (info->cmtd != info->mtds[0])
 			mtd_concat_destroy(info->cmtd);
@@ -88,7 +89,10 @@  static int physmap_flash_remove(struct platform_device *dev)
 	if (physmap_data && physmap_data->exit)
 		physmap_data->exit(dev);
 
-	return 0;
+out:
+	pm_runtime_put(&dev->dev);
+	pm_runtime_disable(&dev->dev);
+	return err;
 }
 
 static void physmap_set_vpp(struct map_info *map, int state)
@@ -484,13 +488,19 @@  static int physmap_flash_probe(struct platform_device *dev)
 		return -EINVAL;
 	}
 
+	pm_runtime_enable(&dev->dev);
+	pm_runtime_get_sync(&dev->dev);
+
 	if (dev->dev.of_node)
 		err = physmap_flash_of_init(dev);
 	else
 		err = physmap_flash_pdata_init(dev);
 
-	if (err)
+	if (err) {
+		pm_runtime_put(&dev->dev);
+		pm_runtime_disable(&dev->dev);
 		return err;
+	}
 
 	for (i = 0; i < info->nmaps; i++) {
 		struct resource *res;