drm/i915/dp: Reset link params on connector disconnect
Subsystems (3)
PATCH 1/1 drm/i915/dp: Reset link params on connector disconnect
We have noticed that when link training fails the panel
sends a long pulse indicating connector disconnect. In this case we need to reset the link parameters instead of continuing
to use the fallback parameters since else this long pulse
by the panel followed by a modeset request which was triggered by the userspace before getting the connector status as disconnected, will
result into a modeset now using lower link rate/lane count values.
--- drivers/gpu/drm/i915/display/intel_dp.c | 28 +++++++++++++++++-------- 1 file changed, 19 insertions(+), 9 deletions(-)
drivers/gpu/drm/i915/display/intel_dp.c +19 -9
@@ -6111,6 +6111,18 @@ intel_dp_unset_edid(struct intel_dp *intel_dp) intel_dp->edid_quirks =0 ; } +static void +intel_dp_reset_link_params (struct intel_dp *intel_dp) +{ +/* Initial max link lane count */ + intel_dp->max_link_lane_count =intel_dp_max_common_lane_count (intel_dp); + +/* Initial max link rate */ + intel_dp->max_link_rate =intel_dp_max_common_rate (intel_dp); + + intel_dp->reset_link_params = false; +} + static int intel_dp_detect(struct drm_connector *connector, struct drm_modeset_acquire_ctx *ctx,
@@ -6139,6 +6151,11 @@ intel_dp_detect(struct drm_connector *connector,memset (&intel_dp->compliance,0 ,sizeof (intel_dp->compliance));memset (intel_dp->dsc_dpcd,0 ,sizeof (intel_dp->dsc_dpcd)); +/*Reset the immutable VRR Capable property */ +drm_connector_set_vrr_capable_property (connector, + false); +intel_dp_reset_link_params (intel_dp); +if (intel_dp->is_mst) { drm_dbg_kms(&dev_priv->drm, "MST device may have disappeared %d vs %d\n",
@@ -6152,15 +6169,8 @@ intel_dp_detect(struct drm_connector *connector, goto out; } -if (intel_dp->reset_link_params) { -/* Initial max link lane count */ - intel_dp->max_link_lane_count =intel_dp_max_common_lane_count (intel_dp); - -/* Initial max link rate */ - intel_dp->max_link_rate =intel_dp_max_common_rate (intel_dp); - - intel_dp->reset_link_params = false; - } +if (intel_dp->reset_link_params) +intel_dp_reset_link_params (intel_dp);intel_dp_print_rates (intel_dp);
Signature · 6 lines
2.19.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Patchwork
CI[Intel-gfx] ✗ Fi.CI.CHECKPATCH: warning for drm/i915/dp: Reset link params on connector disconnect
== Series Details ==
== Series Details ==
Series: drm/i915/dp: Reset link params on connector disconnect
URL : https://patchwork.freedesktop.org/series/77974/
State : warning
== Summary ==
$ dim checkpatch origin/drm-tip 01ef52e81850 drm/i915/dp: Reset link params on connector disconnect -:13: WARNING:COMMIT_LOG_LONG_LINE: Possible unwrapped commit description (prefer a maximum 75 chars per line) #13: by the panel followed by a modeset request which was triggered by the userspace
total: 0 errors, 1 warnings, 0 checks, 46 lines checked
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Patchwork
CI[Intel-gfx] ✗ Fi.CI.BAT: failure for drm/i915/dp: Reset link params on connector disconnect
In reply to Manasi Navare== Series Details ==
== Series Details ==
Series: drm/i915/dp: Reset link params on connector disconnect
URL : https://patchwork.freedesktop.org/series/77974/
State : failure
== Summary ==
CI Bug Log - changes from CI_DRM_8579 -> Patchwork_17862
====================================================
Summary
-------
**FAILURE**
Serious unknown changes coming with Patchwork_17862 absolutely need to be verified manually.
If you think the reported changes have nothing to do with the changes introduced in Patchwork_17862, please notify your bug team to allow them to document this new failure mode, which will reduce false positives in CI.
External URL: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_17862/index.html
Possible new issues
-------------------
Here are the unknown changes that may have been introduced in Patchwork_17862:
### IGT changes ###
#### Possible regressions ####
* igt@kms_chamelium@dp-hpd-fast:
- fi-cml-u2: [PASS][1] -> [INCOMPLETE][2]
[1]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_8579/fi-cml-u2/igt@[email protected]
[2]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_17862/fi-cml-u2/igt@[email protected]
- fi-kbl-7500u: [PASS][3] -> [INCOMPLETE][4]
[3]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_8579/fi-kbl-7500u/igt@[email protected]
[4]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_17862/fi-kbl-7500u/igt@[email protected]
Known issues
------------
Here are the changes found in Patchwork_17862 that come from known issues:
### IGT changes ###
#### Possible fixes ####
* {igt@kms_flip@basic-flip-vs-wf_vblank@b-dvi-d1}:
- fi-bwr-2160: [FAIL][5] ([i915#1928]) -> [PASS][6]
[5]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_8579/fi-bwr-2160/igt@kms_flip@[email protected]
[6]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_17862/fi-bwr-2160/igt@kms_flip@[email protected]
{name}: This element is suppressed. This means it is ignored when computing
the status of the difference (SUCCESS, WARNING, or FAILURE).
[i915#1928]: https://gitlab.freedesktop.org/drm/intel/issues/1928
Participating hosts (51 -> 9)
------------------------------
ERROR: It appears as if the changes made in Patchwork_17862 prevented too many machines from booting.
Missing (42): fi-kbl-soraka fi-bdw-gvtdvm fi-icl-u2 fi-apl-guc fi-snb-2520m fi-icl-y fi-skl-lmem fi-byt-n2820 fi-icl-guc fi-icl-dsi fi-skl-6600u fi-snb-2600 fi-bxt-dsi fi-bdw-5557u fi-cml-s fi-tgl-u fi-bsw-n3050 fi-byt-j1900 fi-glk-dsi fi-hsw-4770 fi-ivb-3770 fi-elk-e7500 fi-kbl-7560u fi-tgl-y fi-bsw-nick fi-skl-6700k2 fi-kbl-r fi-ilk-m540 fi-ehl-1 fi-tgl-dsi fi-skl-guc fi-cfl-8700k fi-hsw-4200u fi-byt-squawks fi-bsw-cyan fi-cfl-guc fi-kbl-guc fi-whl-u fi-kbl-x1275 fi-bsw-kefka fi-byt-clapper fi-bdw-samus
Build changes
-------------
* Linux: CI_DRM_8579 -> Patchwork_17862
CI-20190529: 20190529
CI_DRM_8579: 289eb12c88c49a4ac8d325dc457d8878c7f5bdc0 @ git://anongit.freedesktop.org/gfx-ci/linux
IGT_5694: a9b6c4c74bfddf7d3d2da3be08804fe315945cea @ git://anongit.freedesktop.org/xorg/app/intel-gpu-tools
Patchwork_17862: 01ef52e818506813a16584526c1365017ca7aefa @ git://anongit.freedesktop.org/gfx-ci/linux
== Linux commits ==
01ef52e81850 drm/i915/dp: Reset link params on connector disconnect
== Logs ==
For more details see: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_17862/index.html
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 40 lines · Manasi Navare
We have noticed that when link training fails the panel
sends a long pulse indicating connector disconnect. In this case we need to reset the link parameters instead of continuing
to use the fallback parameters since else this long pulse
by the panel followed by a modeset request which was triggered by the userspace before getting the connector status as disconnected, will
result into a modeset now using lower link rate/lane count values.Cc Jani Nikula <[email protected]>Cc Ville Syrjälä <[email protected]>Signed-off-by Manasi Navare <[email protected]>--- drivers/gpu/drm/i915/display/intel_dp.c | 28 +++++++++++++++++-------- 1 file changed, 19 insertions(+), 9 deletions(-)drivers/gpu/drm/i915/display/intel_dp.c +12
-0@@ -6111,6 +6111,18 @@ intel_dp_unset_edid(struct intel_dp *intel_dp) intel_dp->edid_quirks =0 ; } +static void +intel_dp_reset_link_params (struct intel_dp *intel_dp) +{ +/* Initial max link lane count */ + intel_dp->max_link_lane_count =intel_dp_max_common_lane_count (intel_dp); + +/* Initial max link rate */ + intel_dp->max_link_rate =intel_dp_max_common_rate (intel_dp); + + intel_dp->reset_link_params = false; +} + static int intel_dp_detect(struct drm_connector *connector, struct drm_modeset_acquire_ctx *ctx,
drivers/gpu/drm/i915/display/intel_dp.c intel_dp_detect(struct drm_connector *connector, · line 6158
Quoted 8 lines
drivers/gpu/drm/i915/display/intel_dp.c +5
-0@@ -6139,6 +6151,11 @@ intel_dp_detect(struct drm_connector *connector,memset (&intel_dp->compliance,0 ,sizeof (intel_dp->compliance));memset (intel_dp->dsc_dpcd,0 ,sizeof (intel_dp->dsc_dpcd)); +/*Reset the immutable VRR Capable property */ +drm_connector_set_vrr_capable_property (connector, + false); +intel_dp_reset_link_params (intel_dp); +
Why would we care what those are when the sink is disconnected?
Quoted 23 lines
drivers/gpu/drm/i915/display/intel_dp.c +2
-9if (intel_dp->is_mst) { drm_dbg_kms(&dev_priv->drm, "MST device may have disappeared %d vs %d\n",@@ -6152,15 +6169,8 @@ intel_dp_detect(struct drm_connector *connector, goto out; } -if (intel_dp->reset_link_params) { -/* Initial max link lane count */ - intel_dp->max_link_lane_count =intel_dp_max_common_lane_count (intel_dp); - -/* Initial max link rate */ - intel_dp->max_link_rate =intel_dp_max_common_rate (intel_dp); - - intel_dp->reset_link_params = false; - } +if (intel_dp->reset_link_params) +intel_dp_reset_link_params (intel_dp);intel_dp_print_rates (intel_dp);Signature · 1 line
2.19.1
Signature · 6 lines
Ville Syrjälä Intel _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 44 lines · Ville Syrjälä
Quoted 43 lines · Manasi Navare
We have noticed that when link training fails the panel
sends a long pulse indicating connector disconnect. In this case we need to reset the link parameters instead of continuing
to use the fallback parameters since else this long pulse
by the panel followed by a modeset request which was triggered by the userspace before getting the connector status as disconnected, will
result into a modeset now using lower link rate/lane count values.Cc Jani Nikula <[email protected]>Cc Ville Syrjälä <[email protected]>Signed-off-by Manasi Navare <[email protected]>--- drivers/gpu/drm/i915/display/intel_dp.c | 28 +++++++++++++++++-------- 1 file changed, 19 insertions(+), 9 deletions(-)drivers/gpu/drm/i915/display/intel_dp.c +12
-0@@ -6111,6 +6111,18 @@ intel_dp_unset_edid(struct intel_dp *intel_dp) intel_dp->edid_quirks =0 ; } +static void +intel_dp_reset_link_params (struct intel_dp *intel_dp) +{ +/* Initial max link lane count */ + intel_dp->max_link_lane_count =intel_dp_max_common_lane_count (intel_dp); + +/* Initial max link rate */ + intel_dp->max_link_rate =intel_dp_max_common_rate (intel_dp); + + intel_dp->reset_link_params = false; +} + static int intel_dp_detect(struct drm_connector *connector, struct drm_modeset_acquire_ctx *ctx,@@ -6139,6 +6151,11 @@ intel_dp_detect(struct drm_connector *connector,memset (&intel_dp->compliance,0 ,sizeof (intel_dp->compliance));memset (intel_dp->dsc_dpcd,0 ,sizeof (intel_dp->dsc_dpcd));
Quoted 7 lines
drivers/gpu/drm/i915/display/intel_dp.c intel_dp_detect(struct drm_connector *connector, · line 6158
Quoted 5 lines
drivers/gpu/drm/i915/display/intel_dp.c +5
-0+/*Reset the immutable VRR Capable property */ +drm_connector_set_vrr_capable_property (connector, + false); +intel_dp_reset_link_params (intel_dp); +Why would we care what those are when the sink is disconnected?
We are noticing this happen in case the panel send this long pulse indicating status change to disconnected, while the modeset was already triggered by userspace in this case IGT, so the modeset continues right after i915_hotplug_work_fn so we need to reset all params which fixes the bug mentioned.
Manasi
Quoted 20 lines
Quoted 20 lines
drivers/gpu/drm/i915/display/intel_dp.c +2
-9if (intel_dp->is_mst) { drm_dbg_kms(&dev_priv->drm, "MST device may have disappeared %d vs %d\n",@@ -6152,15 +6169,8 @@ intel_dp_detect(struct drm_connector *connector, goto out; } -if (intel_dp->reset_link_params) { -/* Initial max link lane count */ - intel_dp->max_link_lane_count =intel_dp_max_common_lane_count (intel_dp); - -/* Initial max link rate */ - intel_dp->max_link_rate =intel_dp_max_common_rate (intel_dp); - - intel_dp->reset_link_params = false; - } +if (intel_dp->reset_link_params) +intel_dp_reset_link_params (intel_dp);
Quoted 8 lines
Quoted 4 lines
drivers/gpu/drm/i915/display/intel_dp.c +0
-0intel_dp_print_rates (intel_dp);Signature · 1 line
2.19.1Signature · 2 lines
Ville Syrjälä Intel
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 53 lines · Manasi Navare
Quoted 44 lines · Ville Syrjälä
Quoted 43 lines · Manasi Navare
We have noticed that when link training fails the panel
sends a long pulse indicating connector disconnect. In this case we need to reset the link parameters instead of continuing
to use the fallback parameters since else this long pulse
by the panel followed by a modeset request which was triggered by the userspace before getting the connector status as disconnected, will
result into a modeset now using lower link rate/lane count values.Cc Jani Nikula <[email protected]>Cc Ville Syrjälä <[email protected]>Signed-off-by Manasi Navare <[email protected]>--- drivers/gpu/drm/i915/display/intel_dp.c | 28 +++++++++++++++++-------- 1 file changed, 19 insertions(+), 9 deletions(-)drivers/gpu/drm/i915/display/intel_dp.c +12
-0@@ -6111,6 +6111,18 @@ intel_dp_unset_edid(struct intel_dp *intel_dp) intel_dp->edid_quirks =0 ; } +static void +intel_dp_reset_link_params (struct intel_dp *intel_dp) +{ +/* Initial max link lane count */ + intel_dp->max_link_lane_count =intel_dp_max_common_lane_count (intel_dp); + +/* Initial max link rate */ + intel_dp->max_link_rate =intel_dp_max_common_rate (intel_dp); + + intel_dp->reset_link_params = false; +} + static int intel_dp_detect(struct drm_connector *connector, struct drm_modeset_acquire_ctx *ctx,@@ -6139,6 +6151,11 @@ intel_dp_detect(struct drm_connector *connector,memset (&intel_dp->compliance,0 ,sizeof (intel_dp->compliance));memset (intel_dp->dsc_dpcd,0 ,sizeof (intel_dp->dsc_dpcd));Quoted 7 lines
drivers/gpu/drm/i915/display/intel_dp.c intel_dp_detect(struct drm_connector *connector, · line 6158
Quoted 5 lines
drivers/gpu/drm/i915/display/intel_dp.c +5
-0+/*Reset the immutable VRR Capable property */ +drm_connector_set_vrr_capable_property (connector, + false); +intel_dp_reset_link_params (intel_dp); +Why would we care what those are when the sink is disconnected?
Quoted 4 lines
We are noticing this happen in case the panel send this long pulse indicating status change to disconnected, while the modeset was already triggered by userspace in this case IGT, so the modeset continues right after
i915_hotplug_work_fnso we need to reset all params which fixes the bug mentioned.
Why did the link params get out of whack before hpd in the first place?
Quoted 31 lines
Manasi
Quoted 28 lines
Quoted 24 lines
drivers/gpu/drm/i915/display/intel_dp.c +2
-9if (intel_dp->is_mst) { drm_dbg_kms(&dev_priv->drm, "MST device may have disappeared %d vs %d\n",@@ -6152,15 +6169,8 @@ intel_dp_detect(struct drm_connector *connector, goto out; } -if (intel_dp->reset_link_params) { -/* Initial max link lane count */ - intel_dp->max_link_lane_count =intel_dp_max_common_lane_count (intel_dp); - -/* Initial max link rate */ - intel_dp->max_link_rate =intel_dp_max_common_rate (intel_dp); - - intel_dp->reset_link_params = false; - } +if (intel_dp->reset_link_params) +intel_dp_reset_link_params (intel_dp);intel_dp_print_rates (intel_dp);Signature · 1 line
2.19.1Signature · 2 lines
Ville Syrjälä Intel
Signature · 6 lines
Ville Syrjälä Intel _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 54 lines · Ville Syrjälä
Quoted 53 lines · Manasi Navare
Quoted 51 lines · Ville Syrjälä
Quoted 40 lines · Manasi Navare
We have noticed that when link training fails the panel
sends a long pulse indicating connector disconnect. In this case we need to reset the link parameters instead of continuing
to use the fallback parameters since else this long pulse
by the panel followed by a modeset request which was triggered by the userspace before getting the connector status as disconnected, will
result into a modeset now using lower link rate/lane count values.Cc Jani Nikula <[email protected]>Cc Ville Syrjälä <[email protected]>Signed-off-by Manasi Navare <[email protected]>--- drivers/gpu/drm/i915/display/intel_dp.c | 28 +++++++++++++++++-------- 1 file changed, 19 insertions(+), 9 deletions(-)drivers/gpu/drm/i915/display/intel_dp.c +12
-0@@ -6111,6 +6111,18 @@ intel_dp_unset_edid(struct intel_dp *intel_dp) intel_dp->edid_quirks =0 ; } +static void +intel_dp_reset_link_params (struct intel_dp *intel_dp) +{ +/* Initial max link lane count */ + intel_dp->max_link_lane_count =intel_dp_max_common_lane_count (intel_dp); + +/* Initial max link rate */ + intel_dp->max_link_rate =intel_dp_max_common_rate (intel_dp); + + intel_dp->reset_link_params = false; +} + static int intel_dp_detect(struct drm_connector *connector, struct drm_modeset_acquire_ctx *ctx,drivers/gpu/drm/i915/display/intel_dp.c intel_dp_detect(struct drm_connector *connector, · line 6158
Quoted 8 lines
drivers/gpu/drm/i915/display/intel_dp.c +5
-0@@ -6139,6 +6151,11 @@ intel_dp_detect(struct drm_connector *connector,memset (&intel_dp->compliance,0 ,sizeof (intel_dp->compliance));memset (intel_dp->dsc_dpcd,0 ,sizeof (intel_dp->dsc_dpcd)); +/*Reset the immutable VRR Capable property */ +drm_connector_set_vrr_capable_property (connector, + false); +intel_dp_reset_link_params (intel_dp); +Why would we care what those are when the sink is disconnected?
Quoted 7 lines
Quoted 4 lines
We are noticing this happen in case the panel send this long pulse indicating status change to disconnected, while the modeset was already triggered by userspace in this case IGT, so the modeset continues right after
i915_hotplug_work_fnso we need to reset all params which fixes the bug mentioned.Why did the link params get out of whack before hpd in the first place?
Most of the failures, we see the link training fails due to AUX timeouts and then link params fallback to lower values Then we get this hpd, in this case if we dont reset the param to max values, prev triggered modeset continues with fallback values but since connector probe doesnt happen again through IGT, it tries the same mode with fallback values and return encoder config failure.
So after reseting the params, the modeset happens with original values and that time link training passes. This is seen in all kms_atomic_transitions IGT tests
Manasi
Quoted 23 lines
Quoted 23 lines
Manasi
Quoted 20 lines
Quoted 20 lines
drivers/gpu/drm/i915/display/intel_dp.c +2
-9if (intel_dp->is_mst) { drm_dbg_kms(&dev_priv->drm, "MST device may have disappeared %d vs %d\n",@@ -6152,15 +6169,8 @@ intel_dp_detect(struct drm_connector *connector, goto out; } -if (intel_dp->reset_link_params) { -/* Initial max link lane count */ - intel_dp->max_link_lane_count =intel_dp_max_common_lane_count (intel_dp); - -/* Initial max link rate */ - intel_dp->max_link_rate =intel_dp_max_common_rate (intel_dp); - - intel_dp->reset_link_params = false; - } +if (intel_dp->reset_link_params) +intel_dp_reset_link_params (intel_dp);
Quoted 12 lines
Quoted 8 lines
Quoted 8 lines
Quoted 4 lines
drivers/gpu/drm/i915/display/intel_dp.c +0
-0intel_dp_print_rates (intel_dp);Signature · 1 line
2.19.1Signature · 2 lines
Ville Syrjälä IntelSignature · 2 lines
Ville Syrjälä Intel
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 60 lines · Manasi Navare
Quoted 59 lines · Ville Syrjälä
Quoted 57 lines · Manasi Navare
Quoted 44 lines · Ville Syrjälä
Quoted 43 lines · Manasi Navare
We have noticed that when link training fails the panel
sends a long pulse indicating connector disconnect. In this case we need to reset the link parameters instead of continuing
to use the fallback parameters since else this long pulse
by the panel followed by a modeset request which was triggered by the userspace before getting the connector status as disconnected, will
result into a modeset now using lower link rate/lane count values.Cc Jani Nikula <[email protected]>Cc Ville Syrjälä <[email protected]>Signed-off-by Manasi Navare <[email protected]>--- drivers/gpu/drm/i915/display/intel_dp.c | 28 +++++++++++++++++-------- 1 file changed, 19 insertions(+), 9 deletions(-)drivers/gpu/drm/i915/display/intel_dp.c +12
-0@@ -6111,6 +6111,18 @@ intel_dp_unset_edid(struct intel_dp *intel_dp) intel_dp->edid_quirks =0 ; } +static void +intel_dp_reset_link_params (struct intel_dp *intel_dp) +{ +/* Initial max link lane count */ + intel_dp->max_link_lane_count =intel_dp_max_common_lane_count (intel_dp); + +/* Initial max link rate */ + intel_dp->max_link_rate =intel_dp_max_common_rate (intel_dp); + + intel_dp->reset_link_params = false; +} + static int intel_dp_detect(struct drm_connector *connector, struct drm_modeset_acquire_ctx *ctx,@@ -6139,6 +6151,11 @@ intel_dp_detect(struct drm_connector *connector,memset (&intel_dp->compliance,0 ,sizeof (intel_dp->compliance));memset (intel_dp->dsc_dpcd,0 ,sizeof (intel_dp->dsc_dpcd));Quoted 7 lines
drivers/gpu/drm/i915/display/intel_dp.c intel_dp_detect(struct drm_connector *connector, · line 6158
Quoted 5 lines
drivers/gpu/drm/i915/display/intel_dp.c +5
-0+/*Reset the immutable VRR Capable property */ +drm_connector_set_vrr_capable_property (connector, + false); +intel_dp_reset_link_params (intel_dp); +Why would we care what those are when the sink is disconnected?
We are noticing this happen in case the panel send this long pulse indicating status change to disconnected, while the modeset was already triggered by userspace in this case IGT, so the modeset continues right after
i915_hotplug_work_fnso we need to reset all params which fixes the bug mentioned.
Quoted 7 lines
Quoted 2 lines
Why did the link params get out of whack before hpd in the first place?
Most of the failures, we see the link training fails due to AUX timeouts and then link params fallback to lower values Then we get this hpd, in this case if we dont reset the param to max values, prev triggered modeset continues with fallback values but since connector probe doesnt happen again through IGT, it tries the same mode with fallback values and return encoder config failure.
If the link training failed then clearly the sink didn't like us anymore anyway. So feels like resetting these here is just shifting some race window around a bit, but it could still fail if the sink still doesn't like us.
Would be good if someone was able to figure out why the sink goes bad in the first place.
Quoted 41 lines
So after reseting the params, the modeset happens with original values and that time link training passes. This is seen in all
kms_atomic_transitionsIGT testsManasi
Quoted 35 lines
Quoted 31 lines
Manasi
Quoted 28 lines
Quoted 24 lines
drivers/gpu/drm/i915/display/intel_dp.c +2
-9if (intel_dp->is_mst) { drm_dbg_kms(&dev_priv->drm, "MST device may have disappeared %d vs %d\n",@@ -6152,15 +6169,8 @@ intel_dp_detect(struct drm_connector *connector, goto out; } -if (intel_dp->reset_link_params) { -/* Initial max link lane count */ - intel_dp->max_link_lane_count =intel_dp_max_common_lane_count (intel_dp); - -/* Initial max link rate */ - intel_dp->max_link_rate =intel_dp_max_common_rate (intel_dp); - - intel_dp->reset_link_params = false; - } +if (intel_dp->reset_link_params) +intel_dp_reset_link_params (intel_dp);intel_dp_print_rates (intel_dp);Signature · 1 line
2.19.1Signature · 2 lines
Ville Syrjälä IntelSignature · 2 lines
Ville Syrjälä Intel
Signature · 6 lines
Ville Syrjälä Intel _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 69 lines · Ville Syrjälä
Quoted 60 lines · Manasi Navare
Quoted 59 lines · Ville Syrjälä
Quoted 57 lines · Manasi Navare
Quoted 44 lines · Ville Syrjälä
Quoted 43 lines · Manasi Navare
We have noticed that when link training fails the panel
sends a long pulse indicating connector disconnect. In this case we need to reset the link parameters instead of continuing
to use the fallback parameters since else this long pulse
by the panel followed by a modeset request which was triggered by the userspace before getting the connector status as disconnected, will
result into a modeset now using lower link rate/lane count values.Cc Jani Nikula <[email protected]>Cc Ville Syrjälä <[email protected]>Signed-off-by Manasi Navare <[email protected]>--- drivers/gpu/drm/i915/display/intel_dp.c | 28 +++++++++++++++++-------- 1 file changed, 19 insertions(+), 9 deletions(-)drivers/gpu/drm/i915/display/intel_dp.c +12
-0@@ -6111,6 +6111,18 @@ intel_dp_unset_edid(struct intel_dp *intel_dp) intel_dp->edid_quirks =0 ; } +static void +intel_dp_reset_link_params (struct intel_dp *intel_dp) +{ +/* Initial max link lane count */ + intel_dp->max_link_lane_count =intel_dp_max_common_lane_count (intel_dp); + +/* Initial max link rate */ + intel_dp->max_link_rate =intel_dp_max_common_rate (intel_dp); + + intel_dp->reset_link_params = false; +} + static int intel_dp_detect(struct drm_connector *connector, struct drm_modeset_acquire_ctx *ctx,@@ -6139,6 +6151,11 @@ intel_dp_detect(struct drm_connector *connector,memset (&intel_dp->compliance,0 ,sizeof (intel_dp->compliance));memset (intel_dp->dsc_dpcd,0 ,sizeof (intel_dp->dsc_dpcd));Quoted 7 lines
drivers/gpu/drm/i915/display/intel_dp.c intel_dp_detect(struct drm_connector *connector, · line 6158
Quoted 5 lines
drivers/gpu/drm/i915/display/intel_dp.c +5
-0+/*Reset the immutable VRR Capable property */ +drm_connector_set_vrr_capable_property (connector, + false); +intel_dp_reset_link_params (intel_dp); +Why would we care what those are when the sink is disconnected?
We are noticing this happen in case the panel send this long pulse indicating status change to disconnected, while the modeset was already triggered by userspace in this case IGT, so the modeset continues right after
i915_hotplug_work_fnso we need to reset all params which fixes the bug mentioned.Quoted 7 lines
Quoted 2 lines
Why did the link params get out of whack before hpd in the first place?
Most of the failures, we see the link training fails due to AUX timeouts and then link params fallback to lower values Then we get this hpd, in this case if we dont reset the param to max values, prev triggered modeset continues with fallback values but since connector probe doesnt happen again through IGT, it tries the same mode with fallback values and return encoder config failure.
Quoted 7 lines
If the link training failed then clearly the sink didn't like us anymore anyway. So feels like resetting these here is just shifting some race window around a bit, but it could still fail if the sink still doesn't like us.
Would be good if someone was able to figure out why the sink goes bad in the first place.
Oh, and don't we now have Imre's "weird hpd happened in the middle of the test, don't trust the results" thing in igt?
Quoted 50 lines
Quoted 42 lines
So after reseting the params, the modeset happens with original values and that time link training passes. This is seen in all
kms_atomic_transitionsIGT testsManasi
Quoted 35 lines
Quoted 31 lines
Manasi
Quoted 28 lines
Quoted 24 lines
drivers/gpu/drm/i915/display/intel_dp.c +2
-9if (intel_dp->is_mst) { drm_dbg_kms(&dev_priv->drm, "MST device may have disappeared %d vs %d\n",@@ -6152,15 +6169,8 @@ intel_dp_detect(struct drm_connector *connector, goto out; } -if (intel_dp->reset_link_params) { -/* Initial max link lane count */ - intel_dp->max_link_lane_count =intel_dp_max_common_lane_count (intel_dp); - -/* Initial max link rate */ - intel_dp->max_link_rate =intel_dp_max_common_rate (intel_dp); - - intel_dp->reset_link_params = false; - } +if (intel_dp->reset_link_params) +intel_dp_reset_link_params (intel_dp);intel_dp_print_rates (intel_dp);Signature · 1 line
2.19.1Signature · 2 lines
Ville Syrjälä IntelSignature · 2 lines
Ville Syrjälä IntelSignature · 6 lines
Ville Syrjälä Intel _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Signature · 6 lines
Ville Syrjälä Intel _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 10 lines · Ville Syrjälä
[...]
Quoted 9 lines
Quoted 3 lines
Then we get this hpd, in this case if we dont reset the param to max values, prev triggered modeset continues with fallback values but since connector probe doesnt happen again through IGT, it tries the same mode with fallback values and return encoder config failure.
If the link training failed then clearly the sink didn't like us anymore anyway. So feels like resetting these here is just shifting some race window around a bit, but it could still fail if the sink still doesn't like us.
Quoted 5 lines
Quoted 2 lines
Would be good if someone was able to figure out why the sink goes bad in the first place.
Oh, and don't we now have Imre's "weird hpd happened in the middle of the test, don't trust the results" thing in igt?
An LG and IIyama monitor this happens on disconnect and reconnect after waking from an idle state when modesetting them, not sure if it's the same case.
--Imre
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 14 lines · Imre Deak
Quoted 13 lines · Ville Syrjälä
[...]
Quoted 11 lines
Quoted 3 lines
Then we get this hpd, in this case if we dont reset the param to max values, prev triggered modeset continues with fallback values but since connector probe doesnt happen again through IGT, it tries the same mode with fallback values and return encoder config failure.
If the link training failed then clearly the sink didn't like us anymore anyway. So feels like resetting these here is just shifting some race window around a bit, but it could still fail if the sink still doesn't like us.
Would be good if someone was able to figure out why the sink goes bad in the first place.
Quoted 6 lines
Quoted 2 lines
Oh, and don't we now have Imre's "weird hpd happened in the middle of the test, don't trust the results" thing in igt?
An LG and IIyama monitor this happens on disconnect and reconnect after waking from an idle state when modesetting them, not sure if it's the same case.
Well in this case, it happens just after link training failure due to some AUX timeouts then looks like the panel detects that the link was not enabled and sends this HPD which puts us into connector status changing from connected to disconnected.
But in IGT, we dont get any uevent so we dont reprobe and continue with the next igt_display_commit.
So should we in IGT in kms_atomic_transitions, plane-all-modeset-transitions subtest, should we check the connector status everytime before back to back commit calls?
Like I think in real use case, after a link failure the userspace would get a uevent and respond to it by reprobing a connector, but we dont do that in IGT so these random link failures cause issues like in here.
Manasi
Quoted 2 lines
--Imre
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 14 lines · Imre Deak
Quoted 13 lines · Ville Syrjälä
[...]
Quoted 11 lines
Quoted 3 lines
Then we get this hpd, in this case if we dont reset the param to max values, prev triggered modeset continues with fallback values but since connector probe doesnt happen again through IGT, it tries the same mode with fallback values and return encoder config failure.
If the link training failed then clearly the sink didn't like us anymore anyway. So feels like resetting these here is just shifting some race window around a bit, but it could still fail if the sink still doesn't like us.
Would be good if someone was able to figure out why the sink goes bad in the first place.
Quoted 6 lines
Quoted 2 lines
Oh, and don't we now have Imre's "weird hpd happened in the middle of the test, don't trust the results" thing in igt?
An LG and IIyama monitor this happens on disconnect and reconnect after waking from an idle state when modesetting them, not sure if it's the same case.
Manasi, could you try if a modeset on the monitor after it has been disabled for a while always results in a long HPD pulse a few seconds after the modeset? If so does this also happen when you just modeset in a sequence from one mode to the other not letting the monitor idle? The same monitor should be also tested then with the above sequences on older platforms if it behaves the same on those too.
Quoted 6 lines
--Imre
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 22 lines · Imre Deak
Quoted 14 lines · Imre Deak
Quoted 13 lines · Ville Syrjälä
[...]
Quoted 11 lines
Quoted 3 lines
Then we get this hpd, in this case if we dont reset the param to max values, prev triggered modeset continues with fallback values but since connector probe doesnt happen again through IGT, it tries the same mode with fallback values and return encoder config failure.
If the link training failed then clearly the sink didn't like us anymore anyway. So feels like resetting these here is just shifting some race window around a bit, but it could still fail if the sink still doesn't like us.
Would be good if someone was able to figure out why the sink goes bad in the first place.
Quoted 6 lines
Quoted 2 lines
Oh, and don't we now have Imre's "weird hpd happened in the middle of the test, don't trust the results" thing in igt?
An LG and IIyama monitor this happens on disconnect and reconnect after waking from an idle state when modesetting them, not sure if it's the same case.
Quoted 7 lines
Manasi, could you try if a modeset on the monitor after it has been disabled for a while always results in a long HPD pulse a few seconds after the modeset? If so does this also happen when you just modeset in a sequence from one mode to the other not letting the monitor idle? The same monitor should be also tested then with the above sequences on older platforms if it behaves the same on those too.
This test has been passing on older ICL platforms. But on TGL
we do see these AUX E timeouts once in a while which recover on their
own for the next modeset. Any idea why these spurious AUX timeouts and how I can possibly rootcause why these timeouts are seen only with AUX E?
Manasi
Quoted 6 lines
Quoted 6 lines
--Imre
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 31 lines · Manasi Navare
Quoted 22 lines · Imre Deak
Quoted 14 lines · Imre Deak
Quoted 13 lines · Ville Syrjälä
[...]
Quoted 11 lines
Quoted 3 lines
Then we get this hpd, in this case if we dont reset the param to max values, prev triggered modeset continues with fallback values but since connector probe doesnt happen again through IGT, it tries the same mode with fallback values and return encoder config failure.
If the link training failed then clearly the sink didn't like us anymore anyway. So feels like resetting these here is just shifting some race window around a bit, but it could still fail if the sink still doesn't like us.
Would be good if someone was able to figure out why the sink goes bad in the first place.
Quoted 6 lines
Quoted 2 lines
Oh, and don't we now have Imre's "weird hpd happened in the middle of the test, don't trust the results" thing in igt?
An LG and IIyama monitor this happens on disconnect and reconnect after waking from an idle state when modesetting them, not sure if it's the same case.
Quoted 7 lines
Manasi, could you try if a modeset on the monitor after it has been disabled for a while always results in a long HPD pulse a few seconds after the modeset? If so does this also happen when you just modeset in a sequence from one mode to the other not letting the monitor idle? The same monitor should be also tested then with the above sequences on older platforms if it behaves the same on those too.
Quoted 4 lines
This test has been passing on older ICL platforms. But on TGL we do see these AUX E timeouts once in a while which recover on their own for the next modeset. Any idea why these spurious AUX timeouts and how I can possibly rootcause why these timeouts are seen only with AUX E?
If the monitor is in a disconnected state as you described, then AUX will fail too. So you need to root cause why the monitor gets disconnected. One possibility for that is what I described above. You can't really make a conclusion on a test passing on ICL and not on TGL, the timing can be different. You'd need to check if a disconnect happens due to long HPD pulse when using the same monitor with the sequences I described above, both on TGL and then also on ICL.
Quoted 9 lines
Manasi
Quoted 6 lines
Quoted 6 lines
--Imre
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx