[2/2] mtd: core: set ROOT_DEV for partitions marked as root devices in DT

Message ID 20221019115041.31805-1-zajec5@gmail.com
State New
Headers
Series [1/2] dt-bindings: mtd: partitions: support marking root device partition |

Commit Message

Rafał Miłecki Oct. 19, 2022, 11:50 a.m. UTC
  From: Rafał Miłecki <rafal@milecki.pl>

This adds support for "linux,root-device" binding that is used to mark
root device MTD partition. It's useful for devices using device tree
that don't have bootloader passing root info in cmdline.

Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
---
 drivers/mtd/mtdcore.c | 7 +++++++
 1 file changed, 7 insertions(+)
  

Comments

Rafał Miłecki Oct. 19, 2022, 11:53 a.m. UTC | #1
On 19.10.2022 13:50, Rafał Miłecki wrote:
> From: Rafał Miłecki <rafal@milecki.pl>
> 
> This adds support for "linux,root-device" binding that is used to mark
> root device MTD partition. It's useful for devices using device tree
> that don't have bootloader passing root info in cmdline.
> 
> Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
> ---
>   drivers/mtd/mtdcore.c | 7 +++++++
>   1 file changed, 7 insertions(+)
> 
> diff --git a/drivers/mtd/mtdcore.c b/drivers/mtd/mtdcore.c
> index 07249af4f890..034b06aff660 100644
> --- a/drivers/mtd/mtdcore.c
> +++ b/drivers/mtd/mtdcore.c
> @@ -28,6 +28,7 @@
>   #include <linux/leds.h>
>   #include <linux/debugfs.h>
>   #include <linux/nvmem-provider.h>
> +#include <linux/root_dev.h>
>   
>   #include <linux/mtd/mtd.h>
>   #include <linux/mtd/partitions.h>
> @@ -735,6 +736,12 @@ int add_mtd_device(struct mtd_info *mtd)
>   		not->add(mtd);
>   
>   	mutex_unlock(&mtd_table_mutex);
> +
> +	if (of_find_property(mtd_get_of_node(mtd), "linux,rootfs", NULL)) {

I forgot to re-do "format-patch" after "commit --amend".
s/linux,rootfs/linux,root-device/

I'll fix that in V2 after giving this patchset some time to review.


> +		pr_info("mtd: setting mtd%d (%s) as root device\n", mtd->index, mtd->name);
> +		ROOT_DEV = MKDEV(MTD_BLOCK_MAJOR, mtd->index);
> +	}
> +
>   	/* We _know_ we aren't being removed, because
>   	   our caller is still holding us here. So none
>   	   of this try_ nonsense, and no bitching about it
  
Rob Herring Oct. 21, 2022, 1:50 a.m. UTC | #2
On Wed, Oct 19, 2022 at 01:53:15PM +0200, Rafał Miłecki wrote:
> On 19.10.2022 13:50, Rafał Miłecki wrote:
> > From: Rafał Miłecki <rafal@milecki.pl>
> > 
> > This adds support for "linux,root-device" binding that is used to mark
> > root device MTD partition. It's useful for devices using device tree
> > that don't have bootloader passing root info in cmdline.
> > 
> > Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
> > ---
> >   drivers/mtd/mtdcore.c | 7 +++++++
> >   1 file changed, 7 insertions(+)
> > 
> > diff --git a/drivers/mtd/mtdcore.c b/drivers/mtd/mtdcore.c
> > index 07249af4f890..034b06aff660 100644
> > --- a/drivers/mtd/mtdcore.c
> > +++ b/drivers/mtd/mtdcore.c
> > @@ -28,6 +28,7 @@
> >   #include <linux/leds.h>
> >   #include <linux/debugfs.h>
> >   #include <linux/nvmem-provider.h>
> > +#include <linux/root_dev.h>
> >   #include <linux/mtd/mtd.h>
> >   #include <linux/mtd/partitions.h>
> > @@ -735,6 +736,12 @@ int add_mtd_device(struct mtd_info *mtd)
> >   		not->add(mtd);
> >   	mutex_unlock(&mtd_table_mutex);
> > +
> > +	if (of_find_property(mtd_get_of_node(mtd), "linux,rootfs", NULL)) {
> 
> I forgot to re-do "format-patch" after "commit --amend".
> s/linux,rootfs/linux,root-device/

It's really the filesystem you are tagging, not a device, so I think 
'linux,rootfs' is better.

Rob
  

Patch

diff --git a/drivers/mtd/mtdcore.c b/drivers/mtd/mtdcore.c
index 07249af4f890..034b06aff660 100644
--- a/drivers/mtd/mtdcore.c
+++ b/drivers/mtd/mtdcore.c
@@ -28,6 +28,7 @@ 
 #include <linux/leds.h>
 #include <linux/debugfs.h>
 #include <linux/nvmem-provider.h>
+#include <linux/root_dev.h>
 
 #include <linux/mtd/mtd.h>
 #include <linux/mtd/partitions.h>
@@ -735,6 +736,12 @@  int add_mtd_device(struct mtd_info *mtd)
 		not->add(mtd);
 
 	mutex_unlock(&mtd_table_mutex);
+
+	if (of_find_property(mtd_get_of_node(mtd), "linux,rootfs", NULL)) {
+		pr_info("mtd: setting mtd%d (%s) as root device\n", mtd->index, mtd->name);
+		ROOT_DEV = MKDEV(MTD_BLOCK_MAJOR, mtd->index);
+	}
+
 	/* We _know_ we aren't being removed, because
 	   our caller is still holding us here. So none
 	   of this try_ nonsense, and no bitching about it