Re: [PATCH v4 0/9] drm: bridge: dw-hdmi: Refactor PHY support

[Laurent Pinchart <laurent.pinchart@ideasonboard.com>]

Hi Neil,

On Thursday 02 Mar 2017 12:27:52 Neil Armstrong wrote:
> On 03/01/2017 11:39 PM, Laurent Pinchart wrote:
> > Hello,
> > 
> > This patch series refactors all the PHY handling code in order to allow
> > support of vendor PHYs and Synopsys DWC HDMI 2.0 TX PHYs.
> > 
> > The series starts with a few cleanups and small fixes. Patch 1/9 just
> > removes unused code, patch 2/9 moves the color converter code out of the
> > PHY configure function as it isn't PHY-dependent, and patch 3/9 enables
> > color conversion even for DVI as it is needed to output RGB when the
> > input format is YUV.
> > 
> > The next two patches fix the power down (4/9) and up (5/9) sequences to
> > comply with the HDMI TX PHY specifications. They are the biggest
> > functional changes in the whole set, and have been tested successfully
> > (with the rest of the series) on i.MX6Q and R-Car H3. I'll try to perform
> > tests on RK3288 tomorrow if nobody beats me to it (Neil, that's for you
> > :-)).
> 
> Done !
> 
> Tested on RK3288 on ACT8846 EVB Board and Amlogic S905X P230 Board.
> 
> Tested-by: Neil Armstrong <narmstrong@baylibre.com>

Thank you ! I manage to test it on a remote RK3288 board too today, and the 
PLL lock delay was between 150µs and 450µs.

> > The PLL PHY lock delay has been measured to be between 300µs and 350µs on
> > R-Car H3 and between 400µs and 600µs on i.MX6Q. The PHY power down delay
> > has been measured to be less than 50µs on both platforms, and was often
> > close to instant with power down reported in the first poll iteration. We
> > should thus be more than safe with a 5ms timeout.
> > 
> > Patch 6/9 breaks the PHY operations out. Glue code is then allowed to pass
> > a PHY operations structure to support vendor PHYs. The existing PHY
> > support code is turned into a default Synopsys PHYs implementation for
> > those PHY operations.
> > 
> > Patch 7/9 further refactors the Synopsys PHY configuration function to
> > make
> > it modular, in order to support DWC HDMI 2.0 TX PHYs that have a very
> > different register layout compared to the currently supported PHYs. Glue
> > code is again allowed to provide a custom PHY configuration
> > implementation, with the existing PHY support code turned into the
> > default implementation for all currently supported Synopsys PHYs.
> > 
> > Patch 8/9 is a small cleanup that removes the now unneeded device type for
> > glue code platform data, and patch 9/9 follows by switching the driver to
> > regmap in order to support vendor-specific register access more easily.
> > 
> > Kieran Bingham (2):
> >   drm: bridge: dw-hdmi: Add support for custom PHY configuration
> >   drm: bridge: dw-hdmi: Remove device type from platform data
> > 
> > Laurent Pinchart (5):
> >   drm: bridge: dw-hdmi: Remove unused functions
> >   drm: bridge: dw-hdmi: Move CSC configuration out of PHY code
> >   drm: bridge: dw-hdmi: Fix the PHY power down sequence
> >   drm: bridge: dw-hdmi: Fix the PHY power up sequence
> >   drm: bridge: dw-hdmi: Create PHY operations
> > 
> > Neil Armstrong (2):
> >   drm: bridge: dw-hdmi: Enable CSC even for DVI
> >   drm: bridge: dw-hdmi: Switch to regmap for register access
> >  
> >  drivers/gpu/drm/bridge/dw-hdmi.c            | 467 +++++++++++++---------
> >  drivers/gpu/drm/imx/dw_hdmi-imx.c           |   2 -
> >  drivers/gpu/drm/rockchip/dw_hdmi-rockchip.c |   1 -
> >  include/drm/bridge/dw_hdmi.h                |  33 +-
> >  4 files changed, 304 insertions(+), 199 deletions(-)

-- 
Regards,

Laurent Pinchart