[PATCH 0/6] partitions: GPT: refresh partition tables when necessary

[Ahmad Fatoum]

Hello Sascha,

On 12/3/25 4:19 PM, Sascha Hauer wrote:
> This is a continuation of the patch previously sent separately:
> 
> https://lore.kernel.org/all/20251128111929.2463263-1-s.hauer@pengutronix.de/
> 
> Ahmad's comment to this was that we should write the inactive GPT first
> to make sure there's always a valid GPT on the device. This series
> implements just that. I stumbled upon a bug when we restore the primary
> GPT from the alternate GPT, this is fixed in this series.

I thought with this series now applied, we could wipe the first 2
sectors and it would fall back to the alternate GPT, but this is not the
case because partition_table_read() invokes the partition parsers only
when partition_parser_get_by_filetype() detects the magic in the first
two sectors.

Did you consider adapting it?

Thanks,
Ahmad

> 
> Signed-off-by: Sascha Hauer <s.hauer at pengutronix.de>
> ---
> Sascha Hauer (6):
>       partitions: gpt: pass epd context pointer to find_valid_gpt()
>       partitions: gpt: only write actual ptes size to device
>       partitions: gpt: factor out a function to write primary/alternate GPT
>       partitions: gpt: write inactive GPT first
>       partitions: gpt: fix GPT restauration from alternate GPT
>       partitions: gpt: refresh partition tables when necessary
> 
>  common/partitions/Kconfig |  14 +++
>  common/partitions/efi.c   | 293 +++++++++++++++++++++++++++++++++++-----------
>  2 files changed, 236 insertions(+), 71 deletions(-)
> ---
> base-commit: 500df27eb054f86ce232c5bb30c3a7ea8f771e61
> change-id: 20251203-efi-partition-refresh-a18d630961bc
> 
> Best regards,

-- 
Pengutronix e.K.                  |                             |
Steuerwalder Str. 21              | http://www.pengutronix.de/  |
31137 Hildesheim, Germany         | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686  | Fax:   +49-5121-206917-5555 |