[GIT PULL] move Versatile flash protection to the device tree

Brian Norris computersforpeace at gmail.com
Mon Apr 4 09:36:26 PDT 2016


Hi Linus,

On Mon, Apr 04, 2016 at 10:45:48AM +0200, Linus Walleij wrote:
> Hi ARM SoC folks,
> 
> now I'm tired of waiting for the MTD maintainers' word on this patch
> series. I sent a question about this approach on february 13:
> http://marc.info/?l=linux-arm-kernel&m=145537026603174&w=2
> 
> With subsequent pings on feb 22, mar 5, mar 31, and the cleanup
> missing the merge window.
> 
> As you can see the flash protection patch also caters for the
> RealView flash protection bits.
> 
> Please pull this into the ARM SoC tree for testing in linux-next
> for the v4.7 series. If the MTD people appear again and have issues
> you can always drop the branch.
> 
> Yours,
> Linus Walleij
> 
> The following changes since commit 9735a22799b9214d17d3c231fe377fc852f042e9:
> 
>   Linux 4.6-rc2 (2016-04-03 09:09:40 -0500)
> 
> are available in the git repository at:
> 
>   git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-integrator.git
> tags/versatile-flash-for-arm-soc
> 
> for you to fetch changes up to 91011a7605822cd9c16eabcd1cc11ae31d604bfd:
> 
>   ARM: integrator: move flash registration to device tree (2016-04-04
> 10:33:16 +0200)
> 
> ----------------------------------------------------------------
> This is a set of patches removing the board file code
> for Versatile flash handling and instead moving it over
> to the device tree and a special add-on file.
> 
> ----------------------------------------------------------------
> Linus Walleij (4):
>       mtd: physmap_of: add a hook for Versatile write protection
>       mtd: augment the "arm,versatile-flash" bindings

I don't know that much about this hardware, but it's not clear whether
you properly addressed Russell's comment about the sharing of the VPP
control. The reference counting is done on a per-flash basis, so it
doesn't actually handle multiple flash correctly AFAICT. I also don't
know if (like Russell mentioned) this is practically important to
anything besides just phymap/NOR flash, or if there are other flash
types used in practice (if so, then this series is insufficient).

Regarding your prior questions about this driver: I don't really have
much direction for the driver, as I simply inherited it and have barely
touched it. As your series still looks like a reasonable restructuring,
and it isn't a regression from any existing support:

Acked-by: Brian Norris <computersforpeace at gmail.com>

I don't personally care where the PULL goes.

Brian

>       ARM: versatile: move flash registration to the device tree
>       ARM: integrator: move flash registration to device tree
> 
>  .../devicetree/bindings/mtd/arm-versatile.txt      |  20 +-
>  arch/arm/boot/dts/integrator.dtsi                  |   3 +-
>  arch/arm/boot/dts/versatile-ab.dts                 |   5 +-
>  arch/arm/mach-integrator/integrator_ap.c           |  62 -----
>  arch/arm/mach-integrator/integrator_cp.c           |  51 -----
>  arch/arm/mach-versatile/versatile_dt.c             |  47 ----
>  drivers/mtd/maps/Kconfig                           |  10 +
>  drivers/mtd/maps/Makefile                          |   1 +
>  drivers/mtd/maps/physmap_of.c                      |   6 +
>  drivers/mtd/maps/physmap_of_versatile.c            | 253 +++++++++++++++++++++
>  drivers/mtd/maps/physmap_of_versatile.h            |  16 ++
>  11 files changed, 310 insertions(+), 164 deletions(-)
>  create mode 100644 drivers/mtd/maps/physmap_of_versatile.c
>  create mode 100644 drivers/mtd/maps/physmap_of_versatile.h



More information about the linux-arm-kernel mailing list