| Message ID | 1536072941-888-1-git-send-email-shawn.c.lee@intel.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | drm/i915: optimzie eDP 1.4 config | expand |
Quoting Lee, Shawn C (2018-09-04 15:55:41) > eDP 1.4 introduce a new link rates flexibility and selection. > It provided system specific link rate optimization and power > efficiency. We should keep eDP 1.3 and older version to use > max link rate approach to avoid any side effect. And eDP 1.4 > used the optimization link rate and lane count setting. > > Cc: Matt Atwood <matthew.s.atwood@intel.com> > Signed-off-by: Lee, Shawn C <shawn.c.lee@intel.com> > --- > drivers/gpu/drm/i915/intel_dp.c | 6 ++++-- > 1 file changed, 4 insertions(+), 2 deletions(-) > > diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c > index 436c22de33b6..903d640fe712 100644 > --- a/drivers/gpu/drm/i915/intel_dp.c > +++ b/drivers/gpu/drm/i915/intel_dp.c > @@ -1953,8 +1953,10 @@ intel_dp_compute_link_config(struct intel_encoder *encoder, > * configuration, and typically these values correspond to the > * native resolution of the panel. > */ Having just finished a comment explaining why we override the limits for eDP, the following logic is confusing without at least a mention above. > - limits.min_lane_count = limits.max_lane_count; > - limits.min_clock = limits.max_clock; > + if (intel_dp->edp_dpcd[0] <= DP_EDP_13) { > + limits.min_lane_count = limits.max_lane_count; > + limits.min_clock = limits.max_clock; > + } > } > > intel_dp_adjust_compliance_config(intel_dp, pipe_config, &limits); > -- > 2.7.4 > > _______________________________________________ > Intel-gfx mailing list > Intel-gfx@lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/intel-gfx
On Tue, 04 Sep 2018, Chris Wilson <chris@chris-wilson.co.uk> wrote: > Quoting Lee, Shawn C (2018-09-04 15:55:41) >> eDP 1.4 introduce a new link rates flexibility and selection. >> It provided system specific link rate optimization and power >> efficiency. We should keep eDP 1.3 and older version to use >> max link rate approach to avoid any side effect. And eDP 1.4 >> used the optimization link rate and lane count setting. >> >> Cc: Matt Atwood <matthew.s.atwood@intel.com> >> Signed-off-by: Lee, Shawn C <shawn.c.lee@intel.com> >> --- >> drivers/gpu/drm/i915/intel_dp.c | 6 ++++-- >> 1 file changed, 4 insertions(+), 2 deletions(-) >> >> diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c >> index 436c22de33b6..903d640fe712 100644 >> --- a/drivers/gpu/drm/i915/intel_dp.c >> +++ b/drivers/gpu/drm/i915/intel_dp.c >> @@ -1953,8 +1953,10 @@ intel_dp_compute_link_config(struct intel_encoder *encoder, >> * configuration, and typically these values correspond to the >> * native resolution of the panel. >> */ > > Having just finished a comment explaining why we override the limits for > eDP, the following logic is confusing without at least a mention above. I think the proper approach with the appropriate explanations is [1]. I think it's harmful to start optimizing on eDP 1.4 without also optimizing for fast and narrow link per spec at the same time. BR, Jani. [1] http://patchwork.freedesktop.org/patch/msgid/20180905095321.13843-1-jani.nikula@intel.com > >> - limits.min_lane_count = limits.max_lane_count; >> - limits.min_clock = limits.max_clock; >> + if (intel_dp->edp_dpcd[0] <= DP_EDP_13) { >> + limits.min_lane_count = limits.max_lane_count; >> + limits.min_clock = limits.max_clock; >> + } >> } >> >> intel_dp_adjust_compliance_config(intel_dp, pipe_config, &limits); >> -- >> 2.7.4 >> >> _______________________________________________ >> Intel-gfx mailing list >> Intel-gfx@lists.freedesktop.org >> https://lists.freedesktop.org/mailman/listinfo/intel-gfx > _______________________________________________ > Intel-gfx mailing list > Intel-gfx@lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/intel-gfx
On Tue, 04 Sep 2018, Chris Wilson <chris@chris-wilson.co.uk> wrote: > Quoting Lee, Shawn C (2018-09-04 15:55:41) >> eDP 1.4 introduce a new link rates flexibility and selection. >> It provided system specific link rate optimization and power >> efficiency. We should keep eDP 1.3 and older version to use >> max link rate approach to avoid any side effect. And eDP 1.4 >> used the optimization link rate and lane count setting. >> >> Cc: Matt Atwood <matthew.s.atwood@intel.com> >> Signed-off-by: Lee, Shawn C <shawn.c.lee@intel.com> >> --- >> drivers/gpu/drm/i915/intel_dp.c | 6 ++++-- >> 1 file changed, 4 insertions(+), 2 deletions(-) >> >> diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c >> index 436c22de33b6..903d640fe712 100644 >> --- a/drivers/gpu/drm/i915/intel_dp.c >> +++ b/drivers/gpu/drm/i915/intel_dp.c >> @@ -1953,8 +1953,10 @@ intel_dp_compute_link_config(struct intel_encoder *encoder, >> * configuration, and typically these values correspond to the >> * native resolution of the panel. >> */ > > Having just finished a comment explaining why we override the limits for > eDP, the following logic is confusing without at least a mention above. I think the proper approach with the appropriate explanations is [1]. I think it's harmful to start optimizing on eDP 1.4 without also optimizing for fast and narrow link per spec at the same time. BR, Jani. [1] http://patchwork.freedesktop.org/patch/msgid/20180905095321.13843-1-jani.nikula@intel.com > >> - limits.min_lane_count = limits.max_lane_count; >> - limits.min_clock = limits.max_clock; >> + if (intel_dp->edp_dpcd[0] <= DP_EDP_13) { >> + limits.min_lane_count = limits.max_lane_count; >> + limits.min_clock = limits.max_clock; >> + } >> } >> >> intel_dp_adjust_compliance_config(intel_dp, pipe_config, &limits); >> -- >> 2.7.4 >> >> _______________________________________________ >> Intel-gfx mailing list >> Intel-gfx@lists.freedesktop.org >> https://lists.freedesktop.org/mailman/listinfo/intel-gfx > _______________________________________________ > Intel-gfx mailing list > Intel-gfx@lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/intel-gfx
diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 436c22de33b6..903d640fe712 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -1953,8 +1953,10 @@ intel_dp_compute_link_config(struct intel_encoder *encoder, * configuration, and typically these values correspond to the * native resolution of the panel. */ - limits.min_lane_count = limits.max_lane_count; - limits.min_clock = limits.max_clock; + if (intel_dp->edp_dpcd[0] <= DP_EDP_13) { + limits.min_lane_count = limits.max_lane_count; + limits.min_clock = limits.max_clock; + } } intel_dp_adjust_compliance_config(intel_dp, pipe_config, &limits);
eDP 1.4 introduce a new link rates flexibility and selection. It provided system specific link rate optimization and power efficiency. We should keep eDP 1.3 and older version to use max link rate approach to avoid any side effect. And eDP 1.4 used the optimization link rate and lane count setting. Cc: Matt Atwood <matthew.s.atwood@intel.com> Signed-off-by: Lee, Shawn C <shawn.c.lee@intel.com> --- drivers/gpu/drm/i915/intel_dp.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)