drm/i915/gen9: Only copy WM results for changed pipes to skl_hw
Subsystems (2)
PATCH 1/1 drm/i915/gen9: Only copy WM results for changed pipes to skl_hw
When we write watermark values to the hardware, those values are stored in dev_priv->wm.skl_hw. However with recent watermark changes, the results structure we're copying from only contains valid watermark and DDB values for the pipes that are actually changing; the values for other pipes remain 0. Thus a blind copy of the entire skl_wm_values structure will clobber the values for unchanged pipes...we need to be more selective and only copy over the values for the changing pipes.
This mistake was hidden until recently due to another bug that caused us to erroneously re-calculate watermarks for all active pipes rather than changing pipes. Only when that bug was fixed was the impact of this bug discovered (e.g., modesets failing with "Requested display configuration exceeds system watermark limitations" messages and leaving watermarks non-functional, even ones initiated by intel_fbdev_restore_mode).
--- drivers/gpu/drm/i915/intel_pm.c | 22 ++++++++++++++++++++-- 1 file changed, 20 insertions(+), 2 deletions(-)
drivers/gpu/drm/i915/intel_pm.c +20 -2
@@ -4018,8 +4018,10 @@ static void skl_update_wm(struct drm_crtc *crtc)struct drm_device *dev = crtc->dev;struct drm_i915_private *dev_priv =to_i915 (dev);struct skl_wm_values *results = &dev_priv->wm.skl_results; +struct skl_wm_values *hw_vals = &dev_priv->wm.skl_hw;struct intel_crtc_state *cstate =to_intel_crtc_state (crtc->state);struct skl_pipe_wm *pipe_wm = &cstate->wm.skl.optimal; +int pipe;if ((results->dirty_pipes &drm_crtc_mask (crtc)) ==0 )return ;
@@ -4031,8 +4033,24 @@ static void skl_update_wm(struct drm_crtc *crtc)skl_write_wm_values (dev_priv, results);skl_flush_wm_values (dev_priv, results); -/* store the new configuration */ - dev_priv->wm.skl_hw = *results; +/* +* Store the new configuration (but only for the pipes that have +* changed; the other values weren't recomputed). +*/ +for_each_pipe_masked (dev_priv, pipe, results->dirty_pipes) { + hw_vals->wm_linetime[pipe] = results->wm_linetime[pipe]; + hw_vals->ddb.pipe[pipe] = results->ddb.pipe[pipe]; +memcpy (&hw_vals->ddb.plane[pipe], &results->ddb.plane[pipe], +sizeof (hw_vals->ddb.plane[pipe])); +memcpy (&hw_vals->ddb.y_plane[pipe], &results->ddb.y_plane[pipe], +sizeof (hw_vals->ddb.y_plane[pipe])); +memcpy (&hw_vals->plane[pipe], + &results->plane[pipe], +sizeof (results->plane[pipe])); +memcpy (&hw_vals->plane_trans[pipe], + &results->plane_trans[pipe], +sizeof (results->plane_trans[pipe])); + }mutex_unlock (&dev_priv->wm.wm_mutex); }
Signature · 6 lines
2.1.4 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Patchwork
CI✗ Ro.CI.BAT: warning for drm/i915/gen9: Only copy WM results for changed pipes to skl_hw
== Series Details ==
== Series Details ==
Series: drm/i915/gen9: Only copy WM results for changed pipes to skl_hw
URL : https://patchwork.freedesktop.org/series/9628/
State : warning
== Summary ==
Series 9628v1 drm/i915/gen9: Only copy WM results for changed pipes to skl_hw
http://patchwork.freedesktop.org/api/1.0/series/9628/revisions/1/mbox
Test drv_module_reload_basic:
pass -> DMESG-WARN (ro-byt-n2820)
Test gem_exec_suspend:
Subgroup basic-s3:
dmesg-warn -> PASS (fi-skl-i5-6260u)fi-kbl-qkkr total:237 pass:163 dwarn:29 dfail:0 fail:4 skip:41 fi-skl-i5-6260u total:237 pass:207 dwarn:0 dfail:0 fail:4 skip:26 fi-skl-i7-6700k total:237 pass:193 dwarn:0 dfail:0 fail:4 skip:40 fi-snb-i7-2600 total:237 pass:179 dwarn:0 dfail:0 fail:4 skip:54 ro-bdw-i5-5250u total:229 pass:204 dwarn:1 dfail:1 fail:0 skip:23 ro-bdw-i7-5557U total:229 pass:204 dwarn:2 dfail:1 fail:0 skip:22 ro-bdw-i7-5600u total:229 pass:190 dwarn:0 dfail:1 fail:0 skip:38 ro-byt-n2820 total:229 pass:179 dwarn:1 dfail:1 fail:3 skip:45 ro-hsw-i3-4010u total:229 pass:197 dwarn:0 dfail:1 fail:0 skip:31 ro-hsw-i7-4770r total:229 pass:197 dwarn:0 dfail:1 fail:0 skip:31 ro-ilk-i7-620lm total:229 pass:157 dwarn:0 dfail:1 fail:1 skip:70 ro-ilk1-i5-650 total:224 pass:157 dwarn:0 dfail:1 fail:1 skip:65 ro-ivb-i7-3770 total:229 pass:188 dwarn:0 dfail:1 fail:0 skip:40 ro-skl3-i5-6260u total:229 pass:208 dwarn:1 dfail:1 fail:0 skip:19 ro-snb-i7-2620M total:229 pass:179 dwarn:0 dfail:1 fail:1 skip:48
ro-bsw-n3050 failed to connect after reboot
Results at /archive/results/CI_IGT_test/RO_Patchwork_1450/
9fc25d3 drm-intel-nightly: 2016y-07m-07d-17h-02m-51s UTC integration manifest b816693 drm/i915/gen9: Only copy WM results for changed pipes to skl_hw
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Just posting this so it's on the ML.
NAK on this patch, this causes leftover cursors on SKL. Additionally, there's a noticeable lag every time I move the cursor from one monitor to another, and the flickering from the broken watermarks still happens.
Quoted 68 lines · Matt Roper
When we write watermark values to the hardware, those values are stored in dev_priv->wm.skl_hw. However with recent watermark changes, the results structure we're copying from only contains valid watermark and DDB values for the pipes that are actually changing; the values for other pipes remain 0. Thus a blind copy of the entire
skl_wm_valuesstructure will clobber the values for unchanged pipes...we need to be more selective and only copy over the values for the changing pipes.This mistake was hidden until recently due to another bug that caused us to erroneously re-calculate watermarks for all active pipes rather than changing pipes. Only when that bug was fixed was the impact of this bug discovered (e.g., modesets failing with "Requested display configuration exceeds system watermark limitations" messages and leaving watermarks non-functional, even ones initiated by
intel_fbdev_restore_mode).Cc Maarten Lankhorst <[email protected]>Cc Lyude Paul <[email protected]>Fixes 734fa01f3a17 ("drm/i915/gen9: Calculate watermarks during atomic'check' (v2)")
Fixes 9b6130227495 ("drm/i915/gen9: Re-allocate DDB only for changed pipes")Signed-off-by Matt Roper <[email protected]>---
drivers/gpu/drm/i915/intel_pm.c | 22 ++++++++++++++++++++--1 file changed, 20 insertions(+), 2 deletions(-)drivers/gpu/drm/i915/intel_pm.c +2
-0@@ -4018,8 +4018,10 @@ static void skl_update_wm(struct drm_crtc *crtc) struct drm_device *dev = crtc->dev; struct drm_i915_private *dev_priv = to_i915(dev); struct skl_wm_values *results = &dev_priv->wm.skl_results; +struct skl_wm_values *hw_vals = &dev_priv->wm.skl_hw; struct intel_crtc_state *cstate = to_intel_crtc_state(crtc->state); struct skl_pipe_wm *pipe_wm = &cstate->wm.skl.optimal; +int pipe;
if ((results->dirty_pipes&drm_crtc_mask(crtc)) == 0) return;drivers/gpu/drm/i915/intel_pm.c +0
-0@@ -4031,8 +4033,24 @@ static void skl_update_wm(struct drm_crtc *crtc) skl_write_wm_values(dev_priv, results); skl_flush_wm_values(dev_priv, results);drivers/gpu/drm/i915/intel_pm.c +10
-2-/* store the new configuration */ - dev_priv->wm.skl_hw = *results; +/* +* Store the new configuration (but only for the pipes that have +* changed; the other values weren't recomputed). +*/ +for_each_pipe_masked (dev_priv, pipe, results->dirty_pipes) { + hw_vals->wm_linetime[pipe] = results->wm_linetime[pipe]; + hw_vals->ddb.pipe[pipe] = results->ddb.pipe[pipe]; +memcpy (&hw_vals->ddb.plane[pipe], &results->ddb.plane[pipe], + sizeof(hw_vals->ddb.plane[pipe])); + memcpy(&hw_vals->ddb.y_plane[pipe], &results-Quoted 1 line
ddb.y_plane[pipe],
drivers/gpu/drm/i915/intel_pm.c +8
-0+ sizeof(hw_vals->ddb.y_plane[pipe])); +memcpy (&hw_vals->plane[pipe], + &results->plane[pipe], + sizeof(results->plane[pipe])); +memcpy (&hw_vals->plane_trans[pipe], + &results->plane_trans[pipe], + sizeof(results->plane_trans[pipe])); + }
mutex_unlock(&dev_priv->wm.wm_mutex);
}
Signature · 6 lines
Cheers, Lyude _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx