[U-Boot] [RFC PATCH v2 2/8] disk: part: refactor to prepare GPT over MTD

[Simon Glass]

Hi Patrick,

On 30 November 2016 at 04:01, Patrick Delaunay
<patrick.delaunay73 at gmail.com> wrote:
> From: Patrick Delaunay <patrick.delaunay at st.com>
>
> - udpate prepare_backup_gpt_header():
>   calculate partition_entry_lba and no more assumed last lba + 1
>
> - split part_print_efi : create sub-functions
>   part_print_gpt()
>   part_get_disk_info()
>
> Signed-off-by: Patrick Delaunay <patrick.delaunay at st.com>
> Signed-off-by: Patrick Delaunay <patrick.delaunay73 at gmail.com>
> ---
>
> Changes in v2: None
>
>  disk/part_efi.c | 115 ++++++++++++++++++++++++++++++++------------------------
>  1 file changed, 65 insertions(+), 50 deletions(-)
>

To my mind there is too much going on in this one patch? I think it
should be perhaps 3 patches.

> diff --git a/disk/part_efi.c b/disk/part_efi.c
> index 1924338..a23c8ea 100644
> --- a/disk/part_efi.c
> +++ b/disk/part_efi.c
> @@ -10,6 +10,7 @@
>   *   when CONFIG_SYS_64BIT_LBA is not defined, lbaint_t is 32 bits; this
>   *   limits the maximum size of addressable storage to < 2 Terra Bytes
>   */
> +

Unrelated change?

>  #include <asm/unaligned.h>
>  #include <common.h>
>  #include <command.h>
> @@ -159,11 +160,12 @@ static void prepare_backup_gpt_header(gpt_header *gpt_h)
>         uint64_t val;
>
>         /* recalculate the values for the Backup GPT Header */
> -       val = le64_to_cpu(gpt_h->my_lba);
> -       gpt_h->my_lba = gpt_h->alternate_lba;
> -       gpt_h->alternate_lba = cpu_to_le64(val);
> -       gpt_h->partition_entry_lba =
> -                       cpu_to_le64(le64_to_cpu(gpt_h->last_usable_lba) + 1);
> +       val = le64_to_cpu(gpt_h->alternate_lba);
> +       gpt_h->alternate_lba = gpt_h->my_lba;
> +       gpt_h->my_lba = cpu_to_le64(val);
> +       val -= (le32_to_cpu(gpt_h->num_partition_entries) *
> +               le32_to_cpu(gpt_h->sizeof_partition_entry));
> +       gpt_h->partition_entry_lba = cpu_to_le64(val);

I'm not sure what these changes are for, but they probably need their
own commit.

Regards,
Simon