Re: [Intel-gfx] [PATCH v4 5/7] drm/i915/hdcp: Fill wired_cmd_in structures at a single place

2022-12-27 Thread Kandpal, Suraj
> > On Thu, 22 Dec 2022, Suraj Kandpal wrote: > > Need to fill wired cmd in structures at a single place as they remain > > same for both gsc and mei. > > It's probably a good idea to have them in a single place, but there's really > no > reason for that place to be part of i915 at all. > > No

Re: [Intel-gfx] [PATCH v4 5/7] drm/i915/hdcp: Fill wired_cmd_in structures at a single place

2022-12-27 Thread Teres Alexis, Alan Previn
Just my 2 cents (im not hdcp expert but understand a little on the GSC vs MEI): IIRC, HW architecture wise, MEI and GSC-CS (command streamer) are both merely transport protocols to a backend-service. The backend-service here being HDCP-subsystem within the security firmware (one of several backe

Re: [Intel-gfx] [PATCH v4 5/7] drm/i915/hdcp: Fill wired_cmd_in structures at a single place

2022-12-27 Thread Jani Nikula
On Thu, 22 Dec 2022, Suraj Kandpal wrote: > Need to fill wired cmd in structures at a single place as they remain > same for both gsc and mei. It's probably a good idea to have them in a single place, but there's really no reason for that place to be part of i915 at all. None of the functions ar