[Intel-gfx] [PATCH] drm/i915/adl_p: CDCLK crawl support for ADL
Matt Roper
matthew.d.roper at intel.com
Tue Jun 1 20:20:27 UTC 2021
On Tue, Jun 01, 2021 at 05:52:48PM +0300, Stanislav Lisovskiy wrote:
> From: Gwan-gyeong Mun <gwan-gyeong.mun at intel.com>
Aren't you (Stan) the original author of this patch? It looks like the
authorship got changed accidentally in one of the preparation rebases.
A couple other quick drive-by comments below.
>
> CDCLK crawl feature allows to change CDCLK frequency
> without disabling the actual PLL and doesn't require
> a full modeset.
>
> Cc: Mika Kahola <mika.kahola at intel.com>
> Signed-off-by: Stanislav Lisovskiy <stanislav.lisovskiy at intel.com>
> Signed-off-by: Jani Nikula <jani.nikula at intel.com>
> Signed-off-by: Gwan-gyeong Mun <gwan-gyeong.mun at intel.com>
> Cc: Stanislav Lisovskiy <stanislav.lisovskiy at intel.com>
> ---
> drivers/gpu/drm/i915/display/intel_cdclk.c | 72 +++++++++++++++++++---
> drivers/gpu/drm/i915/i915_reg.h | 2 +
> 2 files changed, 65 insertions(+), 9 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/display/intel_cdclk.c b/drivers/gpu/drm/i915/display/intel_cdclk.c
> index 618a9e1e2b0c..b9abed82328c 100644
> --- a/drivers/gpu/drm/i915/display/intel_cdclk.c
> +++ b/drivers/gpu/drm/i915/display/intel_cdclk.c
> @@ -1548,6 +1548,35 @@ static void cnl_cdclk_pll_enable(struct drm_i915_private *dev_priv, int vco)
> dev_priv->cdclk.hw.vco = vco;
> }
>
> +static bool has_cdclk_crawl(struct drm_i915_private *i915)
> +{
> + return IS_ALDERLAKE_P(i915);
> +}
Would it make sense to make this a feature flag in the device info
structure?
> +
> +static void gen13_cdclk_pll_crawl(struct drm_i915_private *dev_priv, int vco)
Function name prefix should either be "adlp" or "xelpd." Probably
"adlp" in this case since I think this functionality relates more to the
platform itself than the display architecture version.
Matt
> +{
> + int ratio = DIV_ROUND_CLOSEST(vco, dev_priv->cdclk.hw.ref);
> + u32 val;
> +
> + /* Write PLL ratio without disabling */
> + val = CNL_CDCLK_PLL_RATIO(ratio) | BXT_DE_PLL_PLL_ENABLE;
> + intel_de_write(dev_priv, BXT_DE_PLL_ENABLE, val);
> +
> + /* Submit freq change request */
> + val |= BXT_DE_PLL_FREQ_REQ;
> + intel_de_write(dev_priv, BXT_DE_PLL_ENABLE, val);
> +
> + /* Timeout 200us */
> + if (intel_de_wait_for_set(dev_priv, BXT_DE_PLL_ENABLE,
> + BXT_DE_PLL_LOCK | BXT_DE_PLL_FREQ_REQ_ACK, 1))
> + DRM_ERROR("timeout waiting for FREQ change request ack\n");
> +
> + val &= ~BXT_DE_PLL_FREQ_REQ;
> + intel_de_write(dev_priv, BXT_DE_PLL_ENABLE, val);
> +
> + dev_priv->cdclk.hw.vco = vco;
> +}
> +
> static u32 bxt_cdclk_cd2x_pipe(struct drm_i915_private *dev_priv, enum pipe pipe)
> {
> if (DISPLAY_VER(dev_priv) >= 12) {
> @@ -1620,14 +1649,16 @@ static void bxt_set_cdclk(struct drm_i915_private *dev_priv,
> return;
> }
>
> - if (DISPLAY_VER(dev_priv) >= 11 || IS_CANNONLAKE(dev_priv)) {
> + if (has_cdclk_crawl(dev_priv) && dev_priv->cdclk.hw.vco > 0 && vco > 0) {
> + if (dev_priv->cdclk.hw.vco != vco)
> + gen13_cdclk_pll_crawl(dev_priv, vco);
> + } else if (DISPLAY_VER(dev_priv) >= 11 || IS_CANNONLAKE(dev_priv)) {
> if (dev_priv->cdclk.hw.vco != 0 &&
> dev_priv->cdclk.hw.vco != vco)
> cnl_cdclk_pll_disable(dev_priv);
>
> if (dev_priv->cdclk.hw.vco != vco)
> cnl_cdclk_pll_enable(dev_priv, vco);
> -
> } else {
> if (dev_priv->cdclk.hw.vco != 0 &&
> dev_priv->cdclk.hw.vco != vco)
> @@ -1820,6 +1851,28 @@ void intel_cdclk_uninit_hw(struct drm_i915_private *i915)
> skl_cdclk_uninit_hw(i915);
> }
>
> +static bool intel_cdclk_can_crawl(struct drm_i915_private *dev_priv,
> + const struct intel_cdclk_config *a,
> + const struct intel_cdclk_config *b)
> +{
> + int a_div, b_div;
> +
> + if (!has_cdclk_crawl(dev_priv))
> + return false;
> +
> + /*
> + * The vco and cd2x divider will change independently
> + * from each, so we disallow cd2x change when crawling.
> + */
> + a_div = DIV_ROUND_CLOSEST(a->vco, a->cdclk);
> + b_div = DIV_ROUND_CLOSEST(b->vco, b->cdclk);
> +
> + return a->vco != 0 && b->vco != 0 &&
> + a->vco != b->vco &&
> + a_div == b_div &&
> + a->ref == b->ref;
> +}
> +
> /**
> * intel_cdclk_needs_modeset - Determine if changong between the CDCLK
> * configurations requires a modeset on all pipes
> @@ -2475,7 +2528,7 @@ int intel_modeset_calc_cdclk(struct intel_atomic_state *state)
> struct drm_i915_private *dev_priv = to_i915(state->base.dev);
> const struct intel_cdclk_state *old_cdclk_state;
> struct intel_cdclk_state *new_cdclk_state;
> - enum pipe pipe;
> + enum pipe pipe = INVALID_PIPE;
> int ret;
>
> new_cdclk_state = intel_atomic_get_cdclk_state(state);
> @@ -2527,15 +2580,18 @@ int intel_modeset_calc_cdclk(struct intel_atomic_state *state)
>
> if (drm_atomic_crtc_needs_modeset(&crtc_state->uapi))
> pipe = INVALID_PIPE;
> - } else {
> - pipe = INVALID_PIPE;
> }
>
> - if (pipe != INVALID_PIPE) {
> + if (intel_cdclk_can_crawl(dev_priv,
> + &old_cdclk_state->actual,
> + &new_cdclk_state->actual)) {
> + drm_dbg_kms(&dev_priv->drm,
> + "Can change cdclk via crawl\n");
> + } else if (pipe != INVALID_PIPE) {
> new_cdclk_state->pipe = pipe;
>
> drm_dbg_kms(&dev_priv->drm,
> - "Can change cdclk with pipe %c active\n",
> + "Can change cdclk cd2x divider with pipe %c active\n",
> pipe_name(pipe));
> } else if (intel_cdclk_needs_modeset(&old_cdclk_state->actual,
> &new_cdclk_state->actual)) {
> @@ -2544,8 +2600,6 @@ int intel_modeset_calc_cdclk(struct intel_atomic_state *state)
> if (ret)
> return ret;
>
> - new_cdclk_state->pipe = INVALID_PIPE;
> -
> drm_dbg_kms(&dev_priv->drm,
> "Modeset required for cdclk change\n");
> }
> diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
> index 3b58067a873c..1d1176d1799d 100644
> --- a/drivers/gpu/drm/i915/i915_reg.h
> +++ b/drivers/gpu/drm/i915/i915_reg.h
> @@ -11157,6 +11157,8 @@ enum skl_power_gate {
> #define BXT_DE_PLL_ENABLE _MMIO(0x46070)
> #define BXT_DE_PLL_PLL_ENABLE (1 << 31)
> #define BXT_DE_PLL_LOCK (1 << 30)
> +#define BXT_DE_PLL_FREQ_REQ (1 << 23)
> +#define BXT_DE_PLL_FREQ_REQ_ACK (1 << 22)
> #define CNL_CDCLK_PLL_RATIO(x) (x)
> #define CNL_CDCLK_PLL_RATIO_MASK 0xff
>
> --
> 2.24.1.485.gad05a3d8e5
>
--
Matt Roper
Graphics Software Engineer
VTT-OSGC Platform Enablement
Intel Corporation
(916) 356-2795
More information about the Intel-gfx
mailing list