On 05/11/2020 14:02, Tomi Valkeinen wrote: > From: Sebastian Reichel <sebastian.reichel@xxxxxxxxxxxxx> > > This drops the virtual channel logic. Afterwards DSI clients > request their channel number and get the virtual channel with > the same number or -EBUSY if already in use. > > Signed-off-by: Sebastian Reichel <sebastian.reichel@xxxxxxxxxxxxx> > Signed-off-by: Tomi Valkeinen <tomi.valkeinen@xxxxxx> > --- > .../gpu/drm/omapdrm/displays/panel-dsi-cm.c | 11 ++--- > drivers/gpu/drm/omapdrm/dss/dsi.c | 46 ++++--------------- > drivers/gpu/drm/omapdrm/dss/omapdss.h | 4 +- > 3 files changed, 12 insertions(+), 49 deletions(-) A comment about this (and other virtual channel logic changes): When Sebastian sent the series in Feb, I commented that the VC change is not correct. That's still true, and I did not fix it as I'm not sure how to fix it, and however it is fixed has to be fully inside dsi.c (so it doesn't affect e.g. the panel drivers). However, it works just fine. The case where things will break is when we need to use multiple VCs (OMAP DSI's "configuration blocks", not virtual channel ID numbers). Command mode panels don't need this. Video mode panels might need it, but it's not quite clear to me if it's strictly needed, or is it just an optimization. I hope to get a board with video mode panel soonish, or help Nikolaus to debug this. Tomi -- Texas Instruments Finland Oy, Porkkalankatu 22, 00180 Helsinki. Y-tunnus/Business ID: 0615521-4. Kotipaikka/Domicile: Helsinki
