From: Ville Syrjälä <ville.syrjala@xxxxxxxxxxxxxxx> Let's sort out the plane data_rate stuff properly by accounting each color plane independently. And we reuse the same code and approach for the relative data rate (which is used for plane ddb allocation). Currently it's not even obvious if the relative data rate is really correct in all the cases. The normal data rate stuff is much more obvious with the way it's getting computed ahead of time. Also reusing the same code for both means we at least have just the one set of bugs to fix. I was also pondering potentially replacing the relative data rate entirely with the real data rate, but it's not obvious what effect that would have to the watermarks. Smaller planes would certainly get a bigger proportion of ddb allocated to them, which might result in worse power consumption I guess. But at least with data_rate and rel_data_rate being handled almost identically we could easily change the strategy at any time. Ville Syrjälä (13): drm/i915: Drop pointless total_data_rate argument drm/i915: Drop pointless dev_priv argument drm/i915: Extract skl_ddb_entry_init() drm/i915: Introduce skl_plane_ddb_iter drm/i915: Extract skl_allocate_plane_ddb() drm/i915: Extract skl_crtc_calc_dbuf_bw() drm/i915: Tweak plane ddb allocation tracking drm/i915: Split plane data_rate into data_rate+data_rate_y drm/i915: Extract intel_adjusted_rate() drm/i915: Reuse intel_adjusted_rate() for pfit pixel rate adjustment drm/i915: Pre-calculate plane relative data rate drm/i915: Remove total[] and uv_total[] from ddb allocation drm/i915: s/plane_res_b/blocks/ etc. .../gpu/drm/i915/display/intel_atomic_plane.c | 122 ++-- .../gpu/drm/i915/display/intel_atomic_plane.h | 8 +- drivers/gpu/drm/i915/display/intel_bw.c | 81 +-- drivers/gpu/drm/i915/display/intel_display.c | 93 ++- .../drm/i915/display/intel_display_debugfs.c | 4 +- .../drm/i915/display/intel_display_types.h | 22 +- drivers/gpu/drm/i915/intel_pm.c | 629 +++++++----------- 7 files changed, 421 insertions(+), 538 deletions(-) -- 2.26.2 _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/intel-gfx
