Re: [PATCH v4.1 1/2] drm/rockchip: analogix_dp: introduce the pclk for grf

From: Yakir Yang
Date: Tue Jun 28 2016 - 23:35:31 EST


Doug,

On 06/23/2016 01:16 PM, Doug Anderson wrote:
Yakir,

On Wed, Jun 22, 2016 at 6:58 PM, Yakir Yang <ykk@xxxxxxxxxxxxxx> wrote:
For RK3399's GRF module, if we want to operate the graphic related grf
registers, we need to enable the pclk_vio_grf which supply power for VIO
GRF IOs, so it's better to introduce an optional grf clock in driver.

Signed-off-by: Yakir Yang <ykk@xxxxxxxxxxxxxx>
---
Hi all,

This is an external patch for analogix_dp misc cleanup thread [0]
[0]: https://patchwork.kernel.org/patch/9175613/

BR,
- Yakir

Changes in v4.1:
- Fix compiled error, sorry.
"dp->cgfclk" --> 'dp->grfclk'

Changes in v4:
- Check the the error code properly, 'EPROBE_DEFER' should be returned,
'ENOENT' should assign a NULL point to grfclk, other errors should be
regarded as failed. (Tomasz, Doug, reviewed at Google Gerrit)
[https://chromium-review.googlesource.com/#/c/351821/20/drivers/gpu/drm/rockchip/analogix_dp-rockchip.c@249]
- Add the document about optional 'grf' clock (Tomasz, Doug, reviewed at Google Gerrit)
[https://chromium-review.googlesource.com/#/c/351821/]

Changes in v3:
- Add this patch in v3

.../display/rockchip/analogix_dp-rockchip.txt | 6 ++++++
drivers/gpu/drm/rockchip/analogix_dp-rockchip.c | 23 +++++++++++++++++++---
2 files changed, 26 insertions(+), 3 deletions(-)
I probably would have split into two patches so the bindings was its
own patch, but I don't think it's strictly required.

In any case, this seems good to me.

Reviewed-by: Douglas Anderson <dianders@xxxxxxxxxxxx>

This is not big change, so collect them into one patch should be more cleaner, thanks for your reviewed ;)