[Intel-gfx] [RFC 3/3] drm/i915: Compute CMRR and calculate vtotal
Jani Nikula
jani.nikula at linux.intel.com
Wed Nov 15 09:14:09 UTC 2023
On Wed, 15 Nov 2023, Mitul Golani <mitulkumar.ajitkumar.golani at intel.com> wrote:
> Compute Fixed Average Vtotal/CMRR with resepect to
> userspace VRR enablement. Also calculate required
> parameters in case of CMRR is enabled. During
> intel_vrr_compute_config, CMRR is getting enabled
> based on userspace has enabled Adaptive Sync Vtotal
> mode (Legacy VRR) or not.
>
> Signed-off-by: Mitul Golani <mitulkumar.ajitkumar.golani at intel.com>
> ---
> drivers/gpu/drm/i915/display/intel_display.c | 1 +
> .../drm/i915/display/intel_display_device.h | 1 +
> drivers/gpu/drm/i915/display/intel_vrr.c | 81 +++++++++++++++++--
> 3 files changed, 78 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/display/intel_display.c b/drivers/gpu/drm/i915/display/intel_display.c
> index f5a69309e65a..d61790f8ebb4 100644
> --- a/drivers/gpu/drm/i915/display/intel_display.c
> +++ b/drivers/gpu/drm/i915/display/intel_display.c
> @@ -5479,6 +5479,7 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
> PIPE_CONF_CHECK_I(vrr.guardband);
> PIPE_CONF_CHECK_LLI(cmrr.cmrr_m);
> PIPE_CONF_CHECK_LLI(cmrr.cmrr_n);
> + PIPE_CONF_CHECK_BOOL(cmrr.enable);
> }
>
> #undef PIPE_CONF_CHECK_X
> diff --git a/drivers/gpu/drm/i915/display/intel_display_device.h b/drivers/gpu/drm/i915/display/intel_display_device.h
> index 4299cc452e05..66cbc3a6bbe8 100644
> --- a/drivers/gpu/drm/i915/display/intel_display_device.h
> +++ b/drivers/gpu/drm/i915/display/intel_display_device.h
> @@ -68,6 +68,7 @@ struct drm_printer;
> #define HAS_TRANSCODER(i915, trans) ((DISPLAY_RUNTIME_INFO(i915)->cpu_transcoder_mask & \
> BIT(trans)) != 0)
> #define HAS_VRR(i915) (DISPLAY_VER(i915) >= 11)
> +#define HAS_CMRR(i915) (DISPLAY_VER(i915) >= 20)
> #define INTEL_NUM_PIPES(i915) (hweight8(DISPLAY_RUNTIME_INFO(i915)->pipe_mask))
> #define I915_HAS_HOTPLUG(i915) (DISPLAY_INFO(i915)->has_hotplug)
> #define OVERLAY_NEEDS_PHYSICAL(i915) (DISPLAY_INFO(i915)->overlay_needs_physical)
> diff --git a/drivers/gpu/drm/i915/display/intel_vrr.c b/drivers/gpu/drm/i915/display/intel_vrr.c
> index 1e33661062b3..4a056a71b68d 100644
> --- a/drivers/gpu/drm/i915/display/intel_vrr.c
> +++ b/drivers/gpu/drm/i915/display/intel_vrr.c
> @@ -105,6 +105,52 @@ int intel_vrr_vmax_vblank_start(const struct intel_crtc_state *crtc_state)
> return crtc_state->vrr.vmax - intel_vrr_vblank_exit_length(crtc_state);
> }
>
> +static int
> +is_cmrr_frac_required(struct intel_crtc_state *crtc_state)
Predicate functions need to be bool, not int.
> +{
> + int target_refresh_k, actual_refresh_k;
> + struct drm_display_mode *adjusted_mode = &crtc_state->hw.adjusted_mode;
> +
> + target_refresh_k = drm_mode_vrefresh(adjusted_mode) * 1000;
> + actual_refresh_k = DIV_ROUND_UP(adjusted_mode->crtc_clock * 1000,
> + adjusted_mode->crtc_htotal) * 1000;
> + actual_refresh_k /= adjusted_mode->crtc_vtotal;
> +
> + if (actual_refresh_k == target_refresh_k)
> + return false;
> +
> + return true;
> +}
> +
> +static unsigned int
> +cmrr_get_vtotal(struct intel_crtc_state *crtc_state)
> +{
> + unsigned int muliplierM = 1, muliplierN = 1, vtotal;
Please no camel case, and please fix the typo in multiplier.
> + unsigned int actual_refresh_rate, desired_refresh_rate;
> + unsigned long long actual_pixel_rate;
All this unsigned math gives me an uneasy feeling. Please consider using
signed math.
> + struct drm_display_mode *adjusted_mode = &crtc_state->hw.adjusted_mode;
> +
> + actual_refresh_rate = DIV_ROUND_UP(adjusted_mode->crtc_clock * 1000,
> + adjusted_mode->crtc_htotal) * 1000;
> + actual_refresh_rate /= adjusted_mode->crtc_vtotal;
> + desired_refresh_rate = drm_mode_vrefresh(adjusted_mode);
> + actual_pixel_rate = actual_refresh_rate * adjusted_mode->crtc_vtotal;
> + actual_pixel_rate = (actual_pixel_rate * adjusted_mode->crtc_htotal) / 1000;
> +
> + if (is_cmrr_frac_required(crtc_state)) {
> + muliplierM = 1001;
> + muliplierN = 1000;
> + }
> +
> + crtc_state->cmrr.cmrr_n = DIV_ROUND_UP(desired_refresh_rate *
> + adjusted_mode->crtc_htotal * muliplierN, muliplierM) * muliplierN;
> + vtotal = DIV_ROUND_UP(actual_pixel_rate * muliplierN, crtc_state->cmrr.cmrr_n);
> + crtc_state->cmrr.cmrr_m =
> + (actual_pixel_rate * muliplierM) % crtc_state->cmrr.cmrr_n;
You're probably going to need do_div() for the remainder.
> +
> + return vtotal;
> +}
> +
> void
> intel_vrr_compute_config(struct intel_crtc_state *crtc_state,
> struct drm_connector_state *conn_state)
> @@ -149,6 +195,27 @@ intel_vrr_compute_config(struct intel_crtc_state *crtc_state,
>
> crtc_state->vrr.flipline = crtc_state->vrr.vmin + 1;
>
> + /*
> + * When panel is VRR capable and userspace has
> + * not enabled adaptive sync mode then Fixed Average
> + * Vtotal mode should be enabled.
> + */
> + if (crtc_state->uapi.vrr_enabled) {
> + crtc_state->vrr.enable = true;
> + if (HAS_CMRR(i915))
> + crtc_state->cmrr.enable = false;
What would have set this to true?
> + crtc_state->mode_flags |= I915_MODE_FLAG_VRR;
> + } else {
> + crtc_state->vrr.enable = false;
What would have set this to true?
You could simplify this to
if (crtc_state->uapi.vrr_enabled)
...
else if (HAS_CMRR(i915))
...
> + if (HAS_CMRR(i915)) {
> + crtc_state->cmrr.enable = true;
> + crtc_state->vrr.vmax = cmrr_get_vtotal(crtc_state);
> + crtc_state->vrr.vmin = crtc_state->vrr.vmax;
> + crtc_state->vrr.flipline = crtc_state->vrr.vmin;
> + crtc_state->mode_flags |= I915_MODE_FLAG_VRR;
> + }
> + }
> +
> /*
> * For XE_LPD+, we use guardband and pipeline override
> * is deprecated.
> @@ -161,11 +228,6 @@ intel_vrr_compute_config(struct intel_crtc_state *crtc_state,
> min(255, crtc_state->vrr.vmin - adjusted_mode->crtc_vblank_start -
> crtc_state->framestart_delay - 1);
> }
> -
> - if (crtc_state->uapi.vrr_enabled) {
> - crtc_state->vrr.enable = true;
> - crtc_state->mode_flags |= I915_MODE_FLAG_VRR;
> - }
> }
>
> static u32 trans_vrr_ctl(const struct intel_crtc_state *crtc_state)
> @@ -295,6 +357,15 @@ void intel_vrr_get_config(struct intel_crtc_state *crtc_state)
>
> trans_vrr_ctl = intel_de_read(dev_priv, TRANS_VRR_CTL(cpu_transcoder));
>
> + if (HAS_CMRR(dev_priv)) {
> + crtc_state->cmrr.enable = (trans_vrr_ctl & VRR_CTL_CMRR_ENABLE) &&
> + (trans_vrr_ctl & VRR_CTL_VRR_ENABLE);
> + crtc_state->vrr.enable = trans_vrr_ctl & VRR_CTL_VRR_ENABLE &&
> + !(trans_vrr_ctl & VRR_CTL_CMRR_ENABLE);
> + } else {
> + crtc_state->vrr.enable = trans_vrr_ctl & VRR_CTL_VRR_ENABLE;
> + }
> +
> if (crtc_state->cmrr.enable) {
> cmrr_n_hi = intel_de_read(dev_priv, TRANS_CMRR_N_HI(cpu_transcoder));
> cmrr_n_lo = intel_de_read(dev_priv, TRANS_CMRR_N_LO(cpu_transcoder));
--
Jani Nikula, Intel
More information about the Intel-gfx
mailing list