[U-Boot] mx6sabresd: Select the CONFIG_EFI_PARTITION option

From: Fabio Estevam <fabio.estevam@nxp.com>

With fastboot support enabled, it is useful to be able to list
the eMMC EFI partitions, so select the CONFIG_EFI_PARTITION option.

Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
---
Hi Stefano,

It seems you picked the v1 of my fastboot patch instead of v2.

Hence I am sending a new patch that only selects CONFIG_EFI_PARTITION.

 configs/mx6sabresd_defconfig | 1 +
 1 file changed, 1 insertion(+)

Hi Fabio,

On 17/02/2018 22:44, Fabio Estevam wrote:
> From: Fabio Estevam <fabio.estevam@nxp.com>
> 
> With fastboot support enabled, it is useful to be able to list
> the eMMC EFI partitions, so select the CONFIG_EFI_PARTITION option.
> 
> Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
> ---
> Hi Stefano,
> 
> It seems you picked the v1 of my fastboot patch instead of v2.
> 

mmmh...very bad. I have often the possibility to make errors because
patchworks just not follow the version. I see and read patches on my
e-mailer, but I pick them from patchworks with pwclient. Very often I
saw that patchworks is reporting an old version and in most cases I
search for the cirrect version. In this case, I have not seen it.

I really miss a patchwork's feature that declare all previous version
"obsolete" or "changes requested" when a new version is posted.

> Hence I am sending a new patch that only selects CONFIG_EFI_PARTITION.
> 
>  configs/mx6sabresd_defconfig | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/configs/mx6sabresd_defconfig b/configs/mx6sabresd_defconfig
> index feb385d..55709e8 100644
> --- a/configs/mx6sabresd_defconfig
> +++ b/configs/mx6sabresd_defconfig
> @@ -48,6 +48,7 @@ CONFIG_CMD_EXT4=y
>  CONFIG_CMD_EXT4_WRITE=y
>  CONFIG_CMD_FAT=y
>  CONFIG_CMD_FS_GENERIC=y
> +CONFIG_EFI_PARTITION=y
>  CONFIG_ENV_IS_IN_MMC=y
>  CONFIG_SPI_FLASH=y
>  CONFIG_SPI_FLASH_STMICRO=y
> 

Thanks for checking this, I apply it.

Best regards,
Stefano

Hi Fabio,

On Sun, Feb 18, 2018 at 3:14 AM, Fabio Estevam <festevam@gmail.com> wrote:
> From: Fabio Estevam <fabio.estevam@nxp.com>
>
> With fastboot support enabled, it is useful to be able to list
> the eMMC EFI partitions, so select the CONFIG_EFI_PARTITION option.
>
> Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
> ---
> Hi Stefano,
>
> It seems you picked the v1 of my fastboot patch instead of v2.
>
> Hence I am sending a new patch that only selects CONFIG_EFI_PARTITION.
>
>  configs/mx6sabresd_defconfig | 1 +
>  1 file changed, 1 insertion(+)
>
> diff --git a/configs/mx6sabresd_defconfig b/configs/mx6sabresd_defconfig
> index feb385d..55709e8 100644
> --- a/configs/mx6sabresd_defconfig
> +++ b/configs/mx6sabresd_defconfig
> @@ -48,6 +48,7 @@ CONFIG_CMD_EXT4=y
>  CONFIG_CMD_EXT4_WRITE=y
>  CONFIG_CMD_FAT=y
>  CONFIG_CMD_FS_GENERIC=y
> +CONFIG_EFI_PARTITION=y

Do you have a partition table details, I'm trying GPT as below

+#define ROOT_UUID       "69dad710-2ce4-4e3c-b16c-21a1d49abed3"
+#define PARTS_DEFAULT \
+        "name=loader1,start=1k,size=64K,uuid=${uuid_gpt_loader1};" \
+        "name=loader2,start=69K,size=827K,uuid=${uuid_gpt_loader2};" \
+        "name=uEnv,start=896K,size=128K,uuid=${uuid_gpt_uenv};" \
+        "name=rootfs,size=-,uuid="ROOT_UUID

=> gpt write mmc 0 $partitions
Writing GPT: Partition overlap
error!

Look like loader1 start=1k wont working, it 's always starting
0x00000022 offset if we unspecify start but SPL wont start there
right?

Jagan.

Hi Jagan,

On Fri, Mar 9, 2018 at 9:09 AM, Jagan Teki <jagannadh.teki@gmail.com> wrote:

> Do you have a partition table details, I'm trying GPT as below

Here is the partition scheme that Shawn used in his test:
https://patchwork.ozlabs.org/patch/867705/

On Sat, Mar 10, 2018 at 7:16 AM, Fabio Estevam <festevam@gmail.com> wrote:
> Hi Jagan,
>
> On Fri, Mar 9, 2018 at 9:09 AM, Jagan Teki <jagannadh.teki@gmail.com> wrote:
>
>> Do you have a partition table details, I'm trying GPT as below
>
> Here is the partition scheme that Shawn used in his test:
> https://patchwork.ozlabs.org/patch/867705/

I've seen this before, I'm trying to understand here - how can we
write SPL, u-boot? which are started from sector 2 and 8a

I've observed GPT can hold partition tables on these areas.

Disk /dev/sda: 7389184 sectors, 3.5 GiB
Model: UMS disk 0
Sector size (logical/physical): 512/512 bytes
Disk identifier (GUID): E6BC20C3-1526-44AF-BF61-D494E13B957D
Partition table holds up to 128 entries
Main partition table begins at sector 2 and ends at sector 33
First usable sector is 2048, last usable sector is 7389150
Partitions will be aligned on 2048-sector boundaries
Total free space is 7387103 sectors (3.5 GiB)

When I try to write raw through dd, partition table is broken.

Jagan.