On 01/12/2020 02:17, Laurent Pinchart wrote:
Hi Tomi,
Thank you for the patch.
On Tue, Nov 24, 2020 at 02:45:22PM +0200, Tomi Valkeinen wrote:
The "channel" usage in omap dsi driver is super confusing. We have three different "channels":
- DSI virtual channel ID. This is a number from 0 to 3, included in the
packet payload.
- VC. This is a register block in the DSI IP. There are four of those
blocks. A VC is a DSI "pipeline", with defined fifo settings, data source (cpu or dispc), and some other settings. It has no relation to the 1).
Lovely that it's called VC :-)
Yes. There are 4 VC blocks. My guess is that initially the HW designers meant that one HW VC block is used for one DSI channel (DSI supports 4 channels). But at least I cannot see how the HW could be used that way.
- dispc channel. It's the "pipeline" number dispc uses to send pixel
data.
To clean this up use the following names for each of the above: 1) "channel" 2) "vc" 3) "dispc_channel"
- and 2) will still be prone to confusion :-S Would it help to name 2)
hw_vc or something similar ?
The omap dsi driver uses "vc" and "channel" consistently after this patch, and afaics in the DRM framework it's always "channel". So I kind of agree with you but I hope this is enough. Otherwise I'll spend lots of hours solving the conflicts after rename... Well, a new rename patch could be done later on top if needed.
This patch is mostly about renaming things, but as in some places 1) and 2) have gotten mixed up, additional changes were needed to untangle them. This is mostly just adding a new parameter to some functions so that we pass both the vc and the channel.
It would ease review if this could be split in two patches.
Ok. I'll see if I can split this up without a huge amount of work. They are a bit tangled up... But I agree, this is a bit too big patch.
Tomi