From: "Lisovskiy, Stanislav" <stanislav.lisovskiy@intel.com>
To: Ville Syrjala <ville.syrjala@linux.intel.com>
Cc: intel-gfx@lists.freedesktop.org
Subject: Re: [Intel-gfx] [PATCH 3/3] drm/i915: Workaround broken BIOS DBUF configuration on TGL/RKL
Date: Mon, 7 Feb 2022 09:30:48 +0200 [thread overview]
Message-ID: <20220207073048.GC2471@intel.com> (raw)
In-Reply-To: <20220204141818.1900-3-ville.syrjala@linux.intel.com>
On Fri, Feb 04, 2022 at 04:18:18PM +0200, Ville Syrjala wrote:
> From: Ville Syrjälä <ville.syrjala@linux.intel.com>
>
> On TGL/RKL the BIOS likes to use some kind of bogus DBUF layout
> that doesn't match what the spec recommends. With a single active
> pipe that is not going to be a problem, but with multiple pipes
> active skl_commit_modeset_enables() goes into an infinite loop
> since it can't figure out any order in which it can commit the
> pipes without causing DBUF overlaps between the planes.
>
> We'd need some kind of extra DBUF defrag stage in between to
> make the transition possible. But that is clearly way too complex
> a solution, so in the name of simplicity let's just sanitize the
> DBUF state by simply turning off all planes when we detect a
> pipe encroaching on its neighbours' DBUF slices. We only have
> to disable the primary planes as all other planes should have
> already been disabled (if they somehow were enabled) by
> earlier sanitization steps.
>
> And for good measure let's also sanitize in case the DBUF
> allocations of the pipes already seem to overlap each other.
>
> Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/4762
> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
> ---
> drivers/gpu/drm/i915/display/intel_display.c | 1 +
> drivers/gpu/drm/i915/intel_pm.c | 68 ++++++++++++++++++++
> drivers/gpu/drm/i915/intel_pm.h | 1 +
> 3 files changed, 70 insertions(+)
>
> diff --git a/drivers/gpu/drm/i915/display/intel_display.c b/drivers/gpu/drm/i915/display/intel_display.c
> index df347329d90e..7f512f9e9e5c 100644
> --- a/drivers/gpu/drm/i915/display/intel_display.c
> +++ b/drivers/gpu/drm/i915/display/intel_display.c
> @@ -10649,6 +10649,7 @@ intel_modeset_setup_hw_state(struct drm_device *dev,
> vlv_wm_sanitize(dev_priv);
> } else if (DISPLAY_VER(dev_priv) >= 9) {
> skl_wm_get_hw_state(dev_priv);
> + skl_wm_sanitize(dev_priv);
> } else if (HAS_PCH_SPLIT(dev_priv)) {
> ilk_wm_get_hw_state(dev_priv);
> }
> diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
> index 79d61a2935ea..02084652fe3d 100644
> --- a/drivers/gpu/drm/i915/intel_pm.c
> +++ b/drivers/gpu/drm/i915/intel_pm.c
> @@ -6710,6 +6710,74 @@ void skl_wm_get_hw_state(struct drm_i915_private *dev_priv)
> dbuf_state->enabled_slices = dev_priv->dbuf.enabled_slices;
> }
>
> +static bool skl_dbuf_is_misconfigured(struct drm_i915_private *i915)
> +{
> + const struct intel_dbuf_state *dbuf_state =
> + to_intel_dbuf_state(i915->dbuf.obj.state);
> + struct skl_ddb_entry entries[I915_MAX_PIPES] = {};
> + struct intel_crtc *crtc;
> +
> + for_each_intel_crtc(&i915->drm, crtc) {
> + const struct intel_crtc_state *crtc_state =
> + to_intel_crtc_state(crtc->base.state);
> +
> + entries[crtc->pipe] = crtc_state->wm.skl.ddb;
> + }
> +
> + for_each_intel_crtc(&i915->drm, crtc) {
> + const struct intel_crtc_state *crtc_state =
> + to_intel_crtc_state(crtc->base.state);
> + u8 slices;
> +
> + slices = skl_compute_dbuf_slices(crtc, dbuf_state->active_pipes,
> + dbuf_state->joined_mbus);
> + if (dbuf_state->slices[crtc->pipe] & ~slices)
> + return true;
> +
> + if (skl_ddb_allocation_overlaps(&crtc_state->wm.skl.ddb, entries,
> + I915_MAX_PIPES, crtc->pipe))
> + return true;
> + }
> +
> + return false;
> +}
> +
> +void skl_wm_sanitize(struct drm_i915_private *i915)
> +{
> + struct intel_crtc *crtc;
> +
> + /*
> + * On TGL/RKL (at least) the BIOS likes to assign the planes
> + * to the wrong DBUF slices. This will cause an infinite loop
> + * in skl_commit_modeset_enables() as it can't find a way to
> + * transition between the old bogus DBUF layout to the new
> + * proper DBUF layout without DBUF allocation overlaps between
> + * the planes (which cannot be allowed or else the hardware
> + * may hang). If we detect a bogus DBUF layout just turn off
> + * all the planes so that skl_commit_modeset_enables() can
> + * simply ignore them.
> + */
> + if (!skl_dbuf_is_misconfigured(i915))
> + return;
> +
> + drm_dbg_kms(&i915->drm, "BIOS has misprogrammed the DBUF, disabling all planes\n");
> +
> + for_each_intel_crtc(&i915->drm, crtc) {
> + struct intel_plane *plane = to_intel_plane(crtc->base.primary);
> + const struct intel_plane_state *plane_state =
> + to_intel_plane_state(plane->base.state);
> + struct intel_crtc_state *crtc_state =
> + to_intel_crtc_state(crtc->base.state);
> +
> + if (plane_state->uapi.visible)
> + intel_plane_disable_noatomic(crtc, plane);
> +
> + drm_WARN_ON(&i915->drm, crtc_state->active_planes != 0);
> +
> + memset(&crtc_state->wm.skl.ddb, 0, sizeof(crtc_state->wm.skl.ddb));
> + }
> +}
> +
> static void ilk_pipe_wm_get_hw_state(struct intel_crtc *crtc)
> {
> struct drm_device *dev = crtc->base.dev;
> diff --git a/drivers/gpu/drm/i915/intel_pm.h b/drivers/gpu/drm/i915/intel_pm.h
> index ae821e35d5fc..51705151b842 100644
> --- a/drivers/gpu/drm/i915/intel_pm.h
> +++ b/drivers/gpu/drm/i915/intel_pm.h
> @@ -46,6 +46,7 @@ void skl_pipe_wm_get_hw_state(struct intel_crtc *crtc,
> struct skl_pipe_wm *out);
> void g4x_wm_sanitize(struct drm_i915_private *dev_priv);
> void vlv_wm_sanitize(struct drm_i915_private *dev_priv);
> +void skl_wm_sanitize(struct drm_i915_private *dev_priv);
> bool intel_can_enable_sagv(struct drm_i915_private *dev_priv,
> const struct intel_bw_state *bw_state);
> void intel_sagv_pre_plane_update(struct intel_atomic_state *state);
> --
> 2.34.1
>
next prev parent reply other threads:[~2022-02-07 7:30 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-02-04 14:18 [Intel-gfx] [PATCH 1/3] drm/i915: Allow !join_mbus cases for adlp+ dbuf configuration Ville Syrjala
2022-02-04 14:18 ` [Intel-gfx] [PATCH 2/3] drm/i915: Populate pipe dbuf slices more accurately during readout Ville Syrjala
2022-02-07 7:30 ` Lisovskiy, Stanislav
2022-02-04 14:18 ` [Intel-gfx] [PATCH 3/3] drm/i915: Workaround broken BIOS DBUF configuration on TGL/RKL Ville Syrjala
2022-02-07 7:30 ` Lisovskiy, Stanislav [this message]
2022-02-07 9:45 ` Ville Syrjälä
2022-02-04 15:06 ` [Intel-gfx] ✗ Fi.CI.BAT: failure for series starting with [1/3] drm/i915: Allow !join_mbus cases for adlp+ dbuf configuration Patchwork
2022-02-04 16:27 ` [Intel-gfx] ✗ Fi.CI.IGT: " Patchwork
2022-02-07 7:29 ` [Intel-gfx] [PATCH 1/3] " Lisovskiy, Stanislav
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20220207073048.GC2471@intel.com \
--to=stanislav.lisovskiy@intel.com \
--cc=intel-gfx@lists.freedesktop.org \
--cc=ville.syrjala@linux.intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox