Skip to contentSkip to conversation

drm/i915: Use crtc_state->has_psr instead of CAN_PSR for pipe update

PATCH Superseded Tarun Vyas 3 participants 4 messages Jul 9, 2018

Subsystems (2)

v3 of 5 Superseded Latest is v5 →

lore
Share & tools
lore

Tarun Vyas

author
#0 lore raw

PATCH v3 1/1 drm/i915: Use crtc_state->has_psr instead of CAN_PSR for pipe update

In commit "drm/i915: Wait for PSR exit before checking for vblank evasion", the idea was to limit the PSR IDLE checks when PSR is actually supported. While CAN_PSR does do that check, it doesn't applies on a per-crtc basis. crtc_state->has_psr is a more granular check that only applies to pipe(s) that have PSR enabled.

Currently, the driver supports PSR on port A + transcoder eDP, so only pipe A will wait for PSR to go IDLE, as it should, and other pipes should return immediately.

Without the has_psr check, non-PSR pipe_updates (pipe B/C in this case), end up waiting on PSR pipe (pipe A in this case) to exit PSR, which may incur substantial delays for non-PSR pipe updates alongwith the fact the it doesn't makes any sense.

Fixes a608987970b9 ("drm/i915: Wait for PSR exit before checking for

vblank evasion")

v2: Remove unnecessary parantheses, make checkpatch happy.

v3: Move the has_psr check to intel_psr_wait_for_idle and commit
message changes (DK).

---
 drivers/gpu/drm/i915/intel_drv.h    | 2 +-
 drivers/gpu/drm/i915/intel_psr.c    | 5 ++++-
 drivers/gpu/drm/i915/intel_sprite.c | 2 +-
 3 files changed, 6 insertions(+), 3 deletions(-)
drivers/gpu/drm/i915/intel_drv.h +1 -1
index 61e715ddd0d5..3f7fbaeb1ee9 100644
@@ -1923,7 +1923,7 @@ void intel_psr_compute_config(struct intel_dp *intel_dp,
 void intel_psr_irq_control(struct drm_i915_private *dev_priv, bool debug);
 void intel_psr_irq_handler(struct drm_i915_private *dev_priv, u32 psr_iir);
 void intel_psr_short_pulse(struct intel_dp *intel_dp);
-int intel_psr_wait_for_idle(struct drm_i915_private *dev_priv);
+int intel_psr_wait_for_idle(struct drm_i915_private *dev_priv, const struct intel_crtc_state *crtc_state);
 
 /* intel_runtime_pm.c */
 int intel_power_domains_init(struct drm_i915_private *);
drivers/gpu/drm/i915/intel_psr.c +4 -1
index 23acc9ac8d4d..3691c857cc91 100644
@@ -717,11 +717,14 @@ void intel_psr_disable(struct intel_dp *intel_dp,
 	cancel_work_sync(&dev_priv->psr.work);
 }
 
-int intel_psr_wait_for_idle(struct drm_i915_private *dev_priv)
+int intel_psr_wait_for_idle(struct drm_i915_private *dev_priv, const struct intel_crtc_state *crtc_state)
 {
 	i915_reg_t reg;
 	u32 mask;
 
+	if(!crtc_state->has_psr)
+		return 0;
+
 	/*
 	 * The sole user right now is intel_pipe_update_start(),
 	 * which won't race with psr_enable/disable, which is
drivers/gpu/drm/i915/intel_sprite.c +1 -1
index 4990d6e84ddf..fcd96c52cf5b 100644
@@ -118,7 +118,7 @@ void intel_pipe_update_start(const struct intel_crtc_state *new_crtc_state)
 	 * VBL interrupts will start the PSR exit and prevent a PSR
 	 * re-entry as well.
 	 */
-	if (CAN_PSR(dev_priv) && intel_psr_wait_for_idle(dev_priv))
+	if (intel_psr_wait_for_idle(dev_priv, new_crtc_state))
 		DRM_ERROR("PSR idle timed out, atomic update may fail\n");
 
 	local_irq_disable();
Signature · 6 lines
2.13.5

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#1 lore raw

✗ Fi.CI.CHECKPATCH: warning for drm/i915: Use crtc_state->has_psr instead of CAN_PSR for pipe update (rev3)

== Series Details ==

== Series Details ==

Series: drm/i915: Use crtc_state->has_psr instead of CAN_PSR for pipe update (rev3)
URL : https://patchwork.freedesktop.org/series/46104/
State : warning

== Summary ==

$ dim checkpatch origin/drm-tip
fc0aabf8423b drm/i915: Use crtc_state->has_psr instead of CAN_PSR for pipe update
-:41: WARNING:LONG_LINE: line over 100 characters
#41: FILE: drivers/gpu/drm/i915/intel_drv.h:1926:
+int intel_psr_wait_for_idle(struct drm_i915_private *dev_priv, const struct intel_crtc_state *crtc_state);

-:54: WARNING:LONG_LINE: line over 100 characters
#54: FILE: drivers/gpu/drm/i915/intel_psr.c:720:
+int intel_psr_wait_for_idle(struct drm_i915_private *dev_priv, const struct intel_crtc_state *crtc_state)

-:59: ERROR:SPACING: space required before the open parenthesis '(' #59: FILE: drivers/gpu/drm/i915/intel_psr.c:725:
+ if(!crtc_state->has_psr)

total: 1 errors, 2 warnings, 0 checks, 31 lines checked

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#2 lore raw

✓ Fi.CI.BAT: success for drm/i915: Use crtc_state->has_psr instead of CAN_PSR for pipe update (rev3)

In reply to Tarun Vyas
== Series Details ==

== Series Details ==

Series: drm/i915: Use crtc_state->has_psr instead of CAN_PSR for pipe update (rev3)
URL : https://patchwork.freedesktop.org/series/46104/
State : success

== Summary ==

= CI Bug Log - changes from CI_DRM_4456 -> Patchwork_9598 =

== Summary - SUCCESS ==

No regressions found.

External URL: https://patchwork.freedesktop.org/api/1.0/series/46104/revisions/3/mbox/

== Known issues ==

Here are the changes found in Patchwork_9598 that come from known issues:

=== IGT changes ===

    ==== Issues hit ====

    igt@kms_frontbuffer_tracking@basic:
      fi-hsw-peppy:       PASS -> DMESG-FAIL (fdo#106103, fdo#102614)

fdo#102614 https://bugs.freedesktop.org/show_bug.cgi?id=102614
fdo#106103 https://bugs.freedesktop.org/show_bug.cgi?id=106103

== Participating hosts (47 -> 41) ==

Missing (6): fi-ilk-m540 fi-hsw-4200u fi-byt-squawks fi-bsw-cyan fi-ctg-p8600 fi-kbl-8809g

== Build changes ==

* Linux: CI_DRM_4456 -> Patchwork_9598

CI_DRM_4456: bba42f28e4944fa7ac6ea7c2269504877faef851 @ git://anongit.freedesktop.org/gfx-ci/linux
IGT_4544: 764160f214cd916ddb79408b9f28ac0ad2df40e0 @ git://anongit.freedesktop.org/xorg/app/intel-gpu-tools
Patchwork_9598: fc0aabf8423b2237296a5ad6e4b8ecee50cee64b @ git://anongit.freedesktop.org/gfx-ci/linux

== Linux commits ==

fc0aabf8423b drm/i915: Use crtc_state->has_psr instead of CAN_PSR for pipe update

== Logs ==

For more details see: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_9598/issues.html
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#3 lore raw
In reply to Tarun Vyas
Quoted 15 lines · Tarun Vyas

In commit "drm/i915: Wait for PSR exit before checking for vblank evasion", the idea was to limit the PSR IDLE checks when PSR is actually supported. While CAN_PSR does do that check, it doesn't applies on a per-crtc basis. crtc_state->has_psr is a more granular check that only applies to pipe(s) that have PSR enabled.

Currently, the driver supports PSR on port A + transcoder eDP, so only pipe A will wait for PSR to go IDLE, as it should, and other pipes should return immediately.

Without the has_psr check, non-PSR pipe_updates (pipe B/C in this case), end up waiting on PSR pipe (pipe A in this case) to exit PSR, which may incur substantial delays for non-PSR pipe updates alongwith the fact the it doesn't makes any sense.

Quoted 4 lines
Fixes a608987970b9 ("drm/i915: Wait for PSR exit before checking for

vblank evasion")

v2: Remove unnecessary parantheses, make checkpatch happy.

There are checkpatch failures in this revision too.

Quoted 41 lines

v3: Move the has_psr check to intel_psr_wait_for_idle and commit message changes (DK).

---
 drivers/gpu/drm/i915/intel_drv.h    | 2 +-
 drivers/gpu/drm/i915/intel_psr.c    | 5 ++++-
 drivers/gpu/drm/i915/intel_sprite.c | 2 +-

 3 files changed, 6 insertions(+), 3 deletions(-)
drivers/gpu/drm/i915/intel_drv.h +0 -0
index 61e715ddd0d5..3f7fbaeb1ee9 100644
@@ -1923,7 +1923,7 @@ void intel_psr_compute_config(struct intel_dp
 *intel_dp,

 void intel_psr_irq_control(struct drm_i915_private *dev_priv, bool debug);
 void intel_psr_irq_handler(struct drm_i915_private *dev_priv, u32 psr_iir);
 void intel_psr_short_pulse(struct intel_dp *intel_dp);

drivers/gpu/drm/i915/intel_drv.h +1 -1
-int intel_psr_wait_for_idle(struct drm_i915_private *dev_priv);
+int intel_psr_wait_for_idle(struct drm_i915_private *dev_priv, const

struct intel_crtc_state *crtc_state);

 /* intel_runtime_pm.c */
 int intel_power_domains_init(struct drm_i915_private *);

drivers/gpu/drm/i915/intel_psr.c +0 -0
index 23acc9ac8d4d..3691c857cc91 100644
@@ -717,11 +717,14 @@ void intel_psr_disable(struct intel_dp
 *intel_dp,
  	cancel_work_sync(&dev_priv->psr.work);
  }

drivers/gpu/drm/i915/intel_psr.c +1 -1
-int intel_psr_wait_for_idle(struct drm_i915_private *dev_priv)
+int intel_psr_wait_for_idle(struct drm_i915_private *dev_priv, const
Quoted 2 lines

struct intel_crtc_state *crtc_state)
 {

Derive dev_priv from crtc_state and call it new_crtc_state?

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx