[U-Boot] [PATCH v2 0/3] Serial Flash: call spi_flash_free more coherently

[Jagan Teki]

On 3 November 2014 at 19:31, Valentin Longchamp
<valentin.longchamp at keymile.com> wrote:
> Some board require spi_flash_free to be called after all the accesses,
> in order, for instance, to restore the pin multiplexing configuration in
> the case where the SPI pins are multiplexed.

So, for each probe calls you must need to free the flash even-though
the flash is-in-use, looks some different behavior.

Can you pleas elaborate little more.

>
> This patch series tries to enhance this. Patch 1 adds spi_flash_free
> calls to env_sf so that the SPI interface is always "cleaned up" after
> the env read/writes. Patch 2 adds a 'sf release' command that implicitly
> calls spi_flash_free and is thus the pendant of 'sf probe'. Patch 3 uses
> the 'sf command' for the km_arm board scripts.
>
> The whole series had already been sent more than a year ago [1] but it
> was rejected without any feedback. So I send this rebased v2 so that it
> finally gets reviewed and merged.
>
> [1] http://thread.gmane.org/gmane.comp.boot-loaders.u-boot/169723
>
> Changes in v2:
> - Rebased on v2014.10
>
> Valentin Longchamp (3):
>   env_sf: generalize call to spi_flash_free after accesses
>   cmd_sf: add 'release' command

This may not be an optimistic solutions where we couldn't add command
to have generic behavior which does at-least on probe time - IMHO.

>   km_arm: call 'sf release' in the newenv and update scripts
>
>  common/cmd_sf.c             | 13 ++++++++++++-
>  common/env_sf.c             | 34 ++++++++++++++++------------------
>  include/configs/km/km_arm.h |  6 ++++--
>  3 files changed, 32 insertions(+), 21 deletions(-)
>
> --
> 1.8.0.1
>

thanks!
-- 
Jagan.