From: "Nautiyal, Ankit K" <ankit.k.nautiyal@intel.com>
To: "Ville Syrjälä" <ville.syrjala@linux.intel.com>
Cc: <intel-gfx@lists.freedesktop.org>,
<intel-xe@lists.freedesktop.org>, <jani.nikula@linux.intel.com>,
<mitulkumar.ajitkumar.golani@intel.com>
Subject: Re: [PATCH 16/20] drm/i915/display: Use fixed rr timings in intel_set_transcoder_timings_lrr()
Date: Thu, 27 Feb 2025 15:58:58 +0530 [thread overview]
Message-ID: <c9800f86-d3eb-4a58-a597-ca692c2a3b0e@intel.com> (raw)
In-Reply-To: <Z78sqJU5ikVe3mUb@intel.com>
On 2/26/2025 8:30 PM, Ville Syrjälä wrote:
> On Mon, Feb 24, 2025 at 11:47:13AM +0530, Ankit Nautiyal wrote:
>> To have Guardband/Pipeline_full reconfigured seamlessly, move the
>> guardband and pipeline_full checks out from the pure !fastset block in
>> intel_pipe_config_compare().
>> Update the intel_set_transcoder_timings_lrr() function to use
>> fixed refresh rate timings for platforms which always use
>> VRR timing generator.
>>
>> Signed-off-by: Ankit Nautiyal <ankit.k.nautiyal@intel.com>
>> ---
>> drivers/gpu/drm/i915/display/intel_display.c | 43 ++++++++++++++++----
>> drivers/gpu/drm/i915/display/intel_vrr.c | 13 +++++-
>> drivers/gpu/drm/i915/display/intel_vrr.h | 2 +
>> 3 files changed, 48 insertions(+), 10 deletions(-)
>>
>> diff --git a/drivers/gpu/drm/i915/display/intel_display.c b/drivers/gpu/drm/i915/display/intel_display.c
>> index 752ed44d8892..d3c12bc3645c 100644
>> --- a/drivers/gpu/drm/i915/display/intel_display.c
>> +++ b/drivers/gpu/drm/i915/display/intel_display.c
>> @@ -2854,6 +2854,9 @@ static void intel_set_transcoder_timings_lrr(const struct intel_crtc_state *crtc
>> intel_de_write(dev_priv, TRANS_VTOTAL(dev_priv, cpu_transcoder),
>> VACTIVE(crtc_vdisplay - 1) |
>> VTOTAL(crtc_vtotal - 1));
>> +
>> + intel_vrr_set_fixed_rr_timings(crtc_state);
>> + intel_vrr_update_guardband(crtc_state, false);
>> }
>>
>> static void intel_set_pipe_src_size(const struct intel_crtc_state *crtc_state)
>> @@ -5127,17 +5130,36 @@ pipe_config_cx0pll_mismatch(struct drm_printer *p, bool fastset,
>> intel_cx0pll_dump_hw_state(display, b);
>> }
>>
>> -static bool allow_vblank_delay_fastset(const struct intel_crtc_state *old_crtc_state)
>> +static
>> +bool allow_vrr_guardband_fastset(const struct intel_crtc_state *old_crtc_state,
>> + const struct intel_crtc_state *new_crtc_state)
>> +{
>> + struct intel_display *display = to_intel_display(old_crtc_state);
>> +
>> + if (!HAS_VRR(display))
>> + return false;
>> +
>> + if (DISPLAY_VER(display) >= 13)
>> + return new_crtc_state->vrr.guardband != old_crtc_state->vrr.guardband;
>> +
>> + return new_crtc_state->vrr.pipeline_full != old_crtc_state->vrr.pipeline_full;
>> +}
> None of this stuff should be necessary. allow_vblank_delay_fastset()
> only says whether we're allowed the fastset or not. It isn't supposed
> to check if it's necessary or not.
>
>> +
>> +static bool allow_vblank_delay_fastset(const struct intel_crtc_state *old_crtc_state,
>> + const struct intel_crtc_state *new_crtc_state)
>> {
>> struct intel_display *display = to_intel_display(old_crtc_state);
>>
>> /*
>> - * Allow fastboot to fix up vblank delay (handled via LRR
>> - * codepaths), a bit dodgy as the registers aren't
>> + * Allow fastboot to fix up vblank delay, vrr guardband/pipeline_full
>> + * (handled via LRR codepaths), a bit dodgy as the registers aren't
>> * double buffered but seems to be working more or less...
>> */
>> - return HAS_LRR(display) && old_crtc_state->inherited &&
>> - !intel_crtc_has_type(old_crtc_state, INTEL_OUTPUT_DSI);
>> + if (!HAS_LRR(display) || intel_crtc_has_type(old_crtc_state, INTEL_OUTPUT_DSI))
>> + return false;
>> +
>> + return old_crtc_state->inherited ||
>> + allow_vrr_guardband_fastset(old_crtc_state, new_crtc_state);
>> }
>>
>> bool
>> @@ -5273,7 +5295,7 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
>> PIPE_CONF_CHECK_I(name.crtc_hsync_start); \
>> PIPE_CONF_CHECK_I(name.crtc_hsync_end); \
>> PIPE_CONF_CHECK_I(name.crtc_vdisplay); \
>> - if (!fastset || !allow_vblank_delay_fastset(current_config)) \
>> + if (!fastset || !allow_vblank_delay_fastset(current_config, pipe_config)) \
>> PIPE_CONF_CHECK_I(name.crtc_vblank_start); \
>> PIPE_CONF_CHECK_I(name.crtc_vsync_start); \
>> PIPE_CONF_CHECK_I(name.crtc_vsync_end); \
>> @@ -5569,8 +5591,6 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
>> PIPE_CONF_CHECK_I(vrr.vmin);
>> PIPE_CONF_CHECK_I(vrr.vmax);
>> PIPE_CONF_CHECK_I(vrr.flipline);
>> - PIPE_CONF_CHECK_I(vrr.pipeline_full);
>> - PIPE_CONF_CHECK_I(vrr.guardband);
>> PIPE_CONF_CHECK_I(vrr.vsync_start);
>> PIPE_CONF_CHECK_I(vrr.vsync_end);
>> PIPE_CONF_CHECK_LLI(cmrr.cmrr_m);
>> @@ -5578,6 +5598,11 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
>> PIPE_CONF_CHECK_BOOL(cmrr.enable);
>> }
>>
>> + if (!fastset || !allow_vblank_delay_fastset(current_config, pipe_config)) {
>> + PIPE_CONF_CHECK_I(vrr.guardband);
>> + PIPE_CONF_CHECK_I(vrr.pipeline_full);
>> + }
> I'd like to see one patch that just moves these completely out
> of the !fastset block, and a second patch that adds the fastboot
> exception. That way if we do see any problems from the live
> guardband reprogramming we can just revert it.
I have one doubt.
If we remove guardband/pipeline_full from the !fastset block, the
VRR-seamless-switch case will start failing.
This is because the test tries a virtual LRR mode, resulting in a change
in the VRR guardband.
Since its not in fastset block, we will need a full modeset, but the
test expects a seamless switch.
Regards,
Ankit
>
> Hmm, except you haven't changed the vrr_compute_config() yet at
> this point in the series, so changing this would break both fastset
> and state checker, I think. So we might have to move this patch to be
> later in the series.
>
>> +
>> #undef PIPE_CONF_CHECK_X
>> #undef PIPE_CONF_CHECK_I
>> #undef PIPE_CONF_CHECK_LLI
>> @@ -5875,7 +5900,7 @@ static void intel_crtc_check_fastset(const struct intel_crtc_state *old_crtc_sta
>> drm_dbg_kms(&i915->drm, "[CRTC:%d:%s] fastset requirement not met, forcing full modeset\n",
>> crtc->base.base.id, crtc->base.name);
>> } else {
>> - if (allow_vblank_delay_fastset(old_crtc_state))
>> + if (allow_vblank_delay_fastset(old_crtc_state, new_crtc_state))
>> new_crtc_state->update_lrr = true;
>> new_crtc_state->uapi.mode_changed = false;
>> }
>> diff --git a/drivers/gpu/drm/i915/display/intel_vrr.c b/drivers/gpu/drm/i915/display/intel_vrr.c
>> index 11b7edbdb116..e77f5b483b09 100644
>> --- a/drivers/gpu/drm/i915/display/intel_vrr.c
>> +++ b/drivers/gpu/drm/i915/display/intel_vrr.c
>> @@ -284,7 +284,6 @@ int intel_vrr_fixed_rr_flipline(const struct intel_crtc_state *crtc_state)
>> return intel_vrr_fixed_rr_vtotal(crtc_state);
>> }
>>
>> -static
>> void intel_vrr_set_fixed_rr_timings(const struct intel_crtc_state *crtc_state)
>> {
>> struct intel_display *display = to_intel_display(crtc_state);
>> @@ -677,3 +676,15 @@ void intel_vrr_get_config(struct intel_crtc_state *crtc_state)
>> if (crtc_state->vrr.enable)
>> crtc_state->mode_flags |= I915_MODE_FLAG_VRR;
>> }
>> +
>> +void intel_vrr_update_guardband(const struct intel_crtc_state *crtc_state, bool needs_modeset)
>> +{
>> + struct intel_display *display = to_intel_display(crtc_state);
>> + enum transcoder cpu_transcoder = crtc_state->cpu_transcoder;
>> + u32 vrr_ctl = trans_vrr_ctl(crtc_state);
>> +
>> + if (intel_vrr_always_use_vrr_tg(display) && !needs_modeset)
>> + vrr_ctl |= VRR_CTL_VRR_ENABLE;
>> +
>> + intel_de_write(display, TRANS_VRR_CTL(display, cpu_transcoder), vrr_ctl);
>> +}
>> diff --git a/drivers/gpu/drm/i915/display/intel_vrr.h b/drivers/gpu/drm/i915/display/intel_vrr.h
>> index 514822577e8a..e4e9cadcdf9e 100644
>> --- a/drivers/gpu/drm/i915/display/intel_vrr.h
>> +++ b/drivers/gpu/drm/i915/display/intel_vrr.h
>> @@ -35,5 +35,7 @@ int intel_vrr_vmin_vtotal(const struct intel_crtc_state *crtc_state);
>> int intel_vrr_vmax_vblank_start(const struct intel_crtc_state *crtc_state);
>> int intel_vrr_vmin_vblank_start(const struct intel_crtc_state *crtc_state);
>> int intel_vrr_vblank_delay(const struct intel_crtc_state *crtc_state);
>> +void intel_vrr_update_guardband(const struct intel_crtc_state *crtc_state, bool needs_modeset);
>> +void intel_vrr_set_fixed_rr_timings(const struct intel_crtc_state *crtc_state);
>>
>> #endif /* __INTEL_VRR_H__ */
>> --
>> 2.45.2
next prev parent reply other threads:[~2025-02-27 10:29 UTC|newest]
Thread overview: 63+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-02-24 6:16 [PATCH 00/20] Use VRR timing generator for fixed refresh rate modes Ankit Nautiyal
2025-02-24 6:16 ` [PATCH 01/20] drm/i915/vrr: Remove unwanted comment Ankit Nautiyal
2025-02-26 13:06 ` Ville Syrjälä
2025-02-24 6:16 ` [PATCH 02/20] drm/i915:vrr: Separate out functions to compute vmin and vmax Ankit Nautiyal
2025-02-26 12:59 ` Ville Syrjälä
2025-02-27 10:13 ` Nautiyal, Ankit K
2025-02-24 6:17 ` [PATCH 03/20] drm/i915/vrr: Make helpers for cmrr and vrr timings Ankit Nautiyal
2025-02-26 13:07 ` Ville Syrjälä
2025-02-24 6:17 ` [PATCH 04/20] drm/i915/vrr: Disable CMRR Ankit Nautiyal
2025-02-26 13:08 ` Ville Syrjälä
2025-02-24 6:17 ` [PATCH 05/20] drm/i915/vrr: Track vrr.enable only for variable timing Ankit Nautiyal
2025-02-26 13:10 ` Ville Syrjälä
2025-02-27 10:16 ` Nautiyal, Ankit K
2025-02-24 6:17 ` [PATCH 06/20] drm/i915/vrr: Use crtc_vtotal for vmin Ankit Nautiyal
2025-02-26 13:11 ` Ville Syrjälä
2025-02-24 6:17 ` [PATCH 07/20] drm/i915/vrr: Prepare for fixed refresh rate timings Ankit Nautiyal
2025-02-26 13:13 ` Ville Syrjälä
2025-02-24 6:17 ` [PATCH 08/20] drm/i915/display: Enable MSA Ignore Timing PAR only when in not fixed_rr mode Ankit Nautiyal
2025-02-26 13:27 ` Ville Syrjälä
2025-02-27 10:19 ` Nautiyal, Ankit K
2025-02-24 6:17 ` [PATCH 09/20] drm/i915/hdmi: Use VRR Timing generator for HDMI Ankit Nautiyal
2025-02-26 13:30 ` Ville Syrjälä
2025-02-27 10:31 ` Nautiyal, Ankit K
2025-02-24 6:17 ` [PATCH 10/20] drm/i915/dp_mst: Use VRR Timing generator for DP MST Ankit Nautiyal
2025-02-26 13:30 ` Ville Syrjälä
2025-02-24 6:17 ` [PATCH 11/20] drm/i915/display: Disable PSR before disabling VRR Ankit Nautiyal
2025-02-26 13:45 ` Ville Syrjälä
2025-02-27 10:33 ` Nautiyal, Ankit K
2025-02-24 6:17 ` [PATCH 12/20] drm/i915/display: Extend WA 14015406119 for PSR2 Ankit Nautiyal
2025-02-26 13:55 ` Ville Syrjälä
2025-02-27 10:36 ` Nautiyal, Ankit K
2025-02-24 6:17 ` [PATCH 13/20] drm/i915/vrr: Handle joiner with vrr Ankit Nautiyal
2025-02-24 13:47 ` Nautiyal, Ankit K
2025-02-26 15:14 ` Ville Syrjälä
2025-02-27 10:37 ` Nautiyal, Ankit K
2025-02-24 6:17 ` [PATCH 14/20] drm/i915/vrr: Refactor condition for computing vmax and LRR Ankit Nautiyal
2025-02-26 14:01 ` Ville Syrjälä
2025-02-27 10:39 ` Nautiyal, Ankit K
2025-02-24 6:17 ` [PATCH 15/20] drm/i915/vrr: Always set vrr vmax/vmin/flipline in vrr_{enable/disable} Ankit Nautiyal
2025-02-26 14:01 ` Ville Syrjälä
2025-02-24 6:17 ` [PATCH 16/20] drm/i915/display: Use fixed rr timings in intel_set_transcoder_timings_lrr() Ankit Nautiyal
2025-02-26 15:00 ` Ville Syrjälä
2025-02-27 10:28 ` Nautiyal, Ankit K [this message]
2025-02-24 6:17 ` [PATCH 17/20] drm/i915/display: Use fixed_rr timings in modeset sequence Ankit Nautiyal
2025-02-26 15:06 ` Ville Syrjälä
2025-02-26 16:04 ` Ville Syrjälä
2025-02-27 10:44 ` Nautiyal, Ankit K
2025-02-27 10:42 ` Nautiyal, Ankit K
2025-02-24 6:17 ` [PATCH 18/20] drm/i915/vrr: Use fixed timings for platforms that support VRR Ankit Nautiyal
2025-02-26 15:11 ` Ville Syrjälä
2025-02-27 11:01 ` Nautiyal, Ankit K
2025-02-27 21:05 ` Ville Syrjälä
2025-02-28 12:31 ` Nautiyal, Ankit K
2025-02-24 6:17 ` [PATCH 19/20] drm/i915/vrr: Always use VRR timing generator for MTL+ Ankit Nautiyal
2025-02-24 6:17 ` [PATCH 20/20] drm/i915/display: Add fixed_rr to crtc_state dump Ankit Nautiyal
2025-02-24 17:59 ` ✓ CI.Patch_applied: success for Use VRR timing generator for fixed refresh rate modes (rev5) Patchwork
2025-02-24 17:59 ` ✓ CI.checkpatch: " Patchwork
2025-02-24 18:00 ` ✓ CI.KUnit: " Patchwork
2025-02-24 18:17 ` ✓ CI.Build: " Patchwork
2025-02-24 18:19 ` ✓ CI.Hooks: " Patchwork
2025-02-24 18:21 ` ✗ CI.checksparse: warning " Patchwork
2025-02-24 18:39 ` ✓ Xe.CI.BAT: success " Patchwork
2025-02-24 20:03 ` ✗ Xe.CI.Full: failure " Patchwork
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=c9800f86-d3eb-4a58-a597-ca692c2a3b0e@intel.com \
--to=ankit.k.nautiyal@intel.com \
--cc=intel-gfx@lists.freedesktop.org \
--cc=intel-xe@lists.freedesktop.org \
--cc=jani.nikula@linux.intel.com \
--cc=mitulkumar.ajitkumar.golani@intel.com \
--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