[PATCH v2 2/3] dm: prepare rkmtd UCLASS

Simon Glass sjg at chromium.org
Thu Oct 12 17:28:49 CEST 2023


Hi Johan,

On Thu, 12 Oct 2023 at 04:19, Johan Jonker <jbx6244 at gmail.com> wrote:
>
>
>
> On 10/2/23 03:16, Simon Glass wrote:
> > Hi Johan,
> >
> > On Thu, 28 Sept 2023 at 12:51, Johan Jonker <jbx6244 at gmail.com> wrote:
> >>
> >> Prepare a rkmtd UCLASS in use for writing Rockchip boot blocks
> >> in combination with existing userspace tools and rockusb command.
> >>
> >> Signed-off-by: Johan Jonker <jbx6244 at gmail.com>
> >> Reviewed-by: Kever Yang <kever.yang at rock-chips.com>
> >> ---
> >>  disk/part.c                | 4 ++++
> >>  drivers/block/blk-uclass.c | 1 +
> >>  include/dm/uclass-id.h     | 1 +
> >>  3 files changed, 6 insertions(+)
> >>
> >
> > A new uclass should have a sandbox test and a header file or some
> > other docs describing it.
>
> Hi Simon, Mark,
>
> Found by git bisect. This serie worked fine till this patch was applied:
>
> blk: Add bounce buffer support to read/write operations
> https://source.denx.de/u-boot/u-boot/-/commit/75191f75bce45f3b9aff607c88f17778d3805c61
>
> On Rockchip boards CONFIG_BOUNCE_BUFFER is selected by CONFIG_MMC_DW.
> These DW controllers have there own DMA interface.
> However Rockchip boards CONFIG_DM_DMA is not enabled.
> A virtual disk driver doesn't work with that extra buffer.
>
> This patch is there. What route should we take to get rid of it.

This rkmtd driver should depend on !BOUNCE_BUFFER perhaps? Or if the
bounce buffer is needed by rockchip, then perhaps we need to be able
to enable/disable it at runtime for different uclasses / devices?

Regards,
Simon


More information about the U-Boot mailing list