From 17e8fd119f8292f0e96bbddc68bbf5c15305a957 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Mon, 29 Oct 2018 20:34:53 +0200 Subject: [PATCH 01/98] drm/i915: Eliminate the horrendous format check code MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Replace the messy framebuffer format/modifier validation code with a single call to drm_any_plane_has_format(). The code was extremely annoying to maintain as you had to have a lot of platform checks for different formats. The new code requires zero maintenance. v2: Nuke the modifier checks as well since the core does that too now v3: Call drm_any_plane_has_format() from the driver code v4: Rebase Cc: Dhinakaran Pandiyan Signed-off-by: Ville Syrjälä Link: https://patchwork.freedesktop.org/patch/msgid/20181029183453.28541-2-ville.syrjala@linux.intel.com Reviewed-by: Dhinakaran Pandiyan --- drivers/gpu/drm/i915/intel_display.c | 105 ++------------------------- 1 file changed, 8 insertions(+), 97 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 812ec5ae5c7b..e9f4e22b2a4e 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -14451,7 +14451,6 @@ static int intel_framebuffer_init(struct intel_framebuffer *intel_fb, { struct drm_i915_private *dev_priv = to_i915(obj->base.dev); struct drm_framebuffer *fb = &intel_fb->base; - struct drm_format_name_buf format_name; u32 pitch_limit; unsigned int tiling, stride; int ret = -EINVAL; @@ -14482,39 +14481,14 @@ static int intel_framebuffer_init(struct intel_framebuffer *intel_fb, } } - /* Passed in modifier sanity checking. */ - switch (mode_cmd->modifier[0]) { - case I915_FORMAT_MOD_Y_TILED_CCS: - case I915_FORMAT_MOD_Yf_TILED_CCS: - switch (mode_cmd->pixel_format) { - case DRM_FORMAT_XBGR8888: - case DRM_FORMAT_ABGR8888: - case DRM_FORMAT_XRGB8888: - case DRM_FORMAT_ARGB8888: - break; - default: - DRM_DEBUG_KMS("RC supported only with RGB8888 formats\n"); - goto err; - } - /* fall through */ - case I915_FORMAT_MOD_Yf_TILED: - if (mode_cmd->pixel_format == DRM_FORMAT_C8) { - DRM_DEBUG_KMS("Indexed format does not support Yf tiling\n"); - goto err; - } - /* fall through */ - case I915_FORMAT_MOD_Y_TILED: - if (INTEL_GEN(dev_priv) < 9) { - DRM_DEBUG_KMS("Unsupported tiling 0x%llx!\n", - mode_cmd->modifier[0]); - goto err; - } - break; - case DRM_FORMAT_MOD_LINEAR: - case I915_FORMAT_MOD_X_TILED: - break; - default: - DRM_DEBUG_KMS("Unsupported fb modifier 0x%llx!\n", + if (!drm_any_plane_has_format(&dev_priv->drm, + mode_cmd->pixel_format, + mode_cmd->modifier[0])) { + struct drm_format_name_buf format_name; + + DRM_DEBUG_KMS("unsupported pixel format %s / modifier 0x%llx\n", + drm_get_format_name(mode_cmd->pixel_format, + &format_name), mode_cmd->modifier[0]); goto err; } @@ -14549,69 +14523,6 @@ static int intel_framebuffer_init(struct intel_framebuffer *intel_fb, goto err; } - /* Reject formats not supported by any plane early. */ - switch (mode_cmd->pixel_format) { - case DRM_FORMAT_C8: - case DRM_FORMAT_RGB565: - case DRM_FORMAT_XRGB8888: - case DRM_FORMAT_ARGB8888: - break; - case DRM_FORMAT_XRGB1555: - if (INTEL_GEN(dev_priv) > 3) { - DRM_DEBUG_KMS("unsupported pixel format: %s\n", - drm_get_format_name(mode_cmd->pixel_format, &format_name)); - goto err; - } - break; - case DRM_FORMAT_ABGR8888: - if (!IS_VALLEYVIEW(dev_priv) && !IS_CHERRYVIEW(dev_priv) && - INTEL_GEN(dev_priv) < 9) { - DRM_DEBUG_KMS("unsupported pixel format: %s\n", - drm_get_format_name(mode_cmd->pixel_format, &format_name)); - goto err; - } - break; - case DRM_FORMAT_XBGR8888: - case DRM_FORMAT_XRGB2101010: - case DRM_FORMAT_XBGR2101010: - if (INTEL_GEN(dev_priv) < 4) { - DRM_DEBUG_KMS("unsupported pixel format: %s\n", - drm_get_format_name(mode_cmd->pixel_format, &format_name)); - goto err; - } - break; - case DRM_FORMAT_ABGR2101010: - if (!IS_VALLEYVIEW(dev_priv) && !IS_CHERRYVIEW(dev_priv)) { - DRM_DEBUG_KMS("unsupported pixel format: %s\n", - drm_get_format_name(mode_cmd->pixel_format, &format_name)); - goto err; - } - break; - case DRM_FORMAT_YUYV: - case DRM_FORMAT_UYVY: - case DRM_FORMAT_YVYU: - case DRM_FORMAT_VYUY: - if (INTEL_GEN(dev_priv) < 5 && !IS_G4X(dev_priv)) { - DRM_DEBUG_KMS("unsupported pixel format: %s\n", - drm_get_format_name(mode_cmd->pixel_format, &format_name)); - goto err; - } - break; - case DRM_FORMAT_NV12: - if (INTEL_GEN(dev_priv) < 9 || IS_SKYLAKE(dev_priv) || - IS_BROXTON(dev_priv)) { - DRM_DEBUG_KMS("unsupported pixel format: %s\n", - drm_get_format_name(mode_cmd->pixel_format, - &format_name)); - goto err; - } - break; - default: - DRM_DEBUG_KMS("unsupported pixel format: %s\n", - drm_get_format_name(mode_cmd->pixel_format, &format_name)); - goto err; - } - /* FIXME need to adjust LINOFF/TILEOFF accordingly. */ if (mode_cmd->offsets[0] != 0) goto err; From 6fc5d789512f7c1a33cbfd420d68d001e899b0de Mon Sep 17 00:00:00 2001 From: Lyude Paul Date: Tue, 20 Nov 2018 19:37:17 -0500 Subject: [PATCH 02/98] drm/i915: Synchronize hpd work in i915_hpd_storm_ctl_show() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit While trying to add a chamelium test for short HPD IRQs, I ran into issues where a hotplug storm would be triggered, but the point at which it would be reported by the kernel would be after igt actually finished checking i915_hpd_storm_ctl's status. So, fix this by simply synchronizing our IRQ work, dig_port_work, and hotplug_work before printing out the HPD storm status in i915_hpd_storm_ctl_show(). Signed-off-by: Lyude Paul Cc: Ville Syrjälä Reviewed-by: Daniel Vetter Link: https://patchwork.freedesktop.org/patch/msgid/20181121003718.17704-1-lyude@redhat.com --- drivers/gpu/drm/i915/i915_debugfs.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c index 7f455bca528e..c120d87df594 100644 --- a/drivers/gpu/drm/i915/i915_debugfs.c +++ b/drivers/gpu/drm/i915/i915_debugfs.c @@ -4592,6 +4592,13 @@ static int i915_hpd_storm_ctl_show(struct seq_file *m, void *data) struct drm_i915_private *dev_priv = m->private; struct i915_hotplug *hotplug = &dev_priv->hotplug; + /* Synchronize with everything first in case there's been an HPD + * storm, but we haven't finished handling it in the kernel yet + */ + synchronize_irq(dev_priv->drm.irq); + flush_work(&dev_priv->hotplug.dig_port_work); + flush_work(&dev_priv->hotplug.hotplug_work); + seq_printf(m, "Threshold: %d\n", hotplug->hpd_storm_threshold); seq_printf(m, "Detected: %s\n", yesno(delayed_work_pending(&hotplug->reenable_work))); From 2bb06265cfd4e4d00ab9612e6a2b9a788f819afd Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Jos=C3=A9=20Roberto=20de=20Souza?= Date: Wed, 21 Nov 2018 14:54:36 -0800 Subject: [PATCH 03/98] drm/i915: Avoid a full port detection in the first eDP short pulse MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Some eDP panels do not set a valid sink count value and even for the ones that sets is should always be one for eDP, that is why it is not cached in intel_edp_init_dpcd(). But intel_dp_short_pulse() compares the old count with the read one if there is a mistmatch a full port detection will be executed, what was happening in the first short pulse interruption of eDP panels that sets sink count. Instead of just skip the compasison for eDP panels, lets not read the sink count at all for eDP. v2: the previous version of this patch it was caching the sink count in intel_edp_init_dpcd() but I was pointed out by Ville a patch that handled a case of a eDP panel that do not set sink count and as sink count is not used to eDP certification was choosed to just not read it at all. Cc: Dhinakaran Pandiyan Reviewed-by: Ville Syrjälä Signed-off-by: José Roberto de Souza Link: https://patchwork.freedesktop.org/patch/msgid/20181121225441.18785-1-jose.souza@intel.com --- drivers/gpu/drm/i915/intel_dp.c | 44 +++++++++++++++++++-------------- 1 file changed, 26 insertions(+), 18 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 7699f9b7b2d2..a6c654e17955 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -3936,8 +3936,6 @@ intel_edp_init_dpcd(struct intel_dp *intel_dp) static bool intel_dp_get_dpcd(struct intel_dp *intel_dp) { - u8 sink_count; - if (!intel_dp_read_dpcd(intel_dp)) return false; @@ -3947,25 +3945,35 @@ intel_dp_get_dpcd(struct intel_dp *intel_dp) intel_dp_set_common_rates(intel_dp); } - if (drm_dp_dpcd_readb(&intel_dp->aux, DP_SINK_COUNT, &sink_count) <= 0) - return false; - /* - * Sink count can change between short pulse hpd hence - * a member variable in intel_dp will track any changes - * between short pulse interrupts. + * Some eDP panels do not set a valid value for sink count, that is why + * it don't care about read it here and in intel_edp_init_dpcd(). */ - intel_dp->sink_count = DP_GET_SINK_COUNT(sink_count); + if (!intel_dp_is_edp(intel_dp)) { + u8 count; + ssize_t r; - /* - * SINK_COUNT == 0 and DOWNSTREAM_PORT_PRESENT == 1 implies that - * a dongle is present but no display. Unless we require to know - * if a dongle is present or not, we don't need to update - * downstream port information. So, an early return here saves - * time from performing other operations which are not required. - */ - if (!intel_dp_is_edp(intel_dp) && !intel_dp->sink_count) - return false; + r = drm_dp_dpcd_readb(&intel_dp->aux, DP_SINK_COUNT, &count); + if (r < 1) + return false; + + /* + * Sink count can change between short pulse hpd hence + * a member variable in intel_dp will track any changes + * between short pulse interrupts. + */ + intel_dp->sink_count = DP_GET_SINK_COUNT(count); + + /* + * SINK_COUNT == 0 and DOWNSTREAM_PORT_PRESENT == 1 implies that + * a dongle is present but no display. Unless we require to know + * if a dongle is present or not, we don't need to update + * downstream port information. So, an early return here saves + * time from performing other operations which are not required. + */ + if (!intel_dp->sink_count) + return false; + } if (!drm_dp_is_branch(intel_dp->dpcd)) return true; /* native DP sink */ From 2f8e7ea974c164c6d2f7761b7406a28fbf0f20c9 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Jos=C3=A9=20Roberto=20de=20Souza?= Date: Wed, 21 Nov 2018 14:54:37 -0800 Subject: [PATCH 04/98] drm/i915: Check PSR errors instead of retrain while PSR is enabled MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit When a PSR error happens sink sets the PSR error register and also set the link status to a error status. So in the short pulse handling it was returning earlier and doing a full detection and attempting to retrain but it fails as PSR HW is in change of the main-link. Just call intel_psr_short_pulse() before intel_dp_needs_link_retrain() is not the right fix as intel_dp_needs_link_retrain() would return true and trigger a full detection while PSR HW is still in change of main-link. Check for PSR active is also not safe as it could be inactive due a frontbuffer invalidate and still doing the PSR exit sequence. v3: added comment in intel_dp_needs_link_retrain() Cc: Dhinakaran Pandiyan Reviewed-by: Rodrigo Vivi Signed-off-by: José Roberto de Souza Link: https://patchwork.freedesktop.org/patch/msgid/20181121225441.18785-2-jose.souza@intel.com --- drivers/gpu/drm/i915/intel_dp.c | 11 +++++++++++ drivers/gpu/drm/i915/intel_drv.h | 1 + drivers/gpu/drm/i915/intel_psr.c | 15 +++++++++++++++ 3 files changed, 27 insertions(+) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index a6c654e17955..70ae3d57316b 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -4383,6 +4383,17 @@ intel_dp_needs_link_retrain(struct intel_dp *intel_dp) if (!intel_dp->link_trained) return false; + /* + * While PSR source HW is enabled, it will control main-link sending + * frames, enabling and disabling it so trying to do a retrain will fail + * as the link would or not be on or it could mix training patterns + * and frame data at the same time causing retrain to fail. + * Also when exiting PSR, HW will retrain the link anyways fixing + * any link status error. + */ + if (intel_psr_enabled(intel_dp)) + return false; + if (!intel_dp_get_link_status(intel_dp, link_status)) return false; diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index a7d9ac912125..a62d77b76291 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -2047,6 +2047,7 @@ 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(const struct intel_crtc_state *new_crtc_state, u32 *out_value); +bool intel_psr_enabled(struct intel_dp *intel_dp); /* intel_quirks.c */ void intel_init_quirks(struct drm_i915_private *dev_priv); diff --git a/drivers/gpu/drm/i915/intel_psr.c b/drivers/gpu/drm/i915/intel_psr.c index 54fa17a5596a..ebb255f230b7 100644 --- a/drivers/gpu/drm/i915/intel_psr.c +++ b/drivers/gpu/drm/i915/intel_psr.c @@ -1147,3 +1147,18 @@ void intel_psr_short_pulse(struct intel_dp *intel_dp) exit: mutex_unlock(&psr->lock); } + +bool intel_psr_enabled(struct intel_dp *intel_dp) +{ + struct drm_i915_private *dev_priv = dp_to_i915(intel_dp); + bool ret; + + if (!CAN_PSR(dev_priv) || !intel_dp_is_edp(intel_dp)) + return false; + + mutex_lock(&dev_priv->psr.lock); + ret = (dev_priv->psr.dp == intel_dp && dev_priv->psr.enabled); + mutex_unlock(&dev_priv->psr.lock); + + return ret; +} From 50a12d8fc9a0d687d9ef4ac9cc16d177a2529937 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Jos=C3=A9=20Roberto=20de=20Souza?= Date: Wed, 21 Nov 2018 14:54:38 -0800 Subject: [PATCH 05/98] drm/i915: Do not enable PSR in the next modeset after a error MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit When we detect a error and disable PSR, it is kept disabled until the next modeset but as the sink already show signs that it do not properly work with PSR lets disabled it for good to avoid any additional flickering. Cc: Dhinakaran Pandiyan Reviewed-by: Rodrigo Vivi Signed-off-by: José Roberto de Souza Link: https://patchwork.freedesktop.org/patch/msgid/20181121225441.18785-3-jose.souza@intel.com --- drivers/gpu/drm/i915/i915_drv.h | 1 + drivers/gpu/drm/i915/intel_psr.c | 10 +++++++++- 2 files changed, 10 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 4064e49dbf70..97cbcb7b016a 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -504,6 +504,7 @@ struct i915_psr { u8 sink_sync_latency; ktime_t last_entry_attempt; ktime_t last_exit; + bool sink_not_reliable; }; enum intel_pch { diff --git a/drivers/gpu/drm/i915/intel_psr.c b/drivers/gpu/drm/i915/intel_psr.c index ebb255f230b7..ab527f9a5436 100644 --- a/drivers/gpu/drm/i915/intel_psr.c +++ b/drivers/gpu/drm/i915/intel_psr.c @@ -527,6 +527,11 @@ void intel_psr_compute_config(struct intel_dp *intel_dp, return; } + if (dev_priv->psr.sink_not_reliable) { + DRM_DEBUG_KMS("PSR sink implementation is not reliable\n"); + return; + } + if (IS_HASWELL(dev_priv) && I915_READ(HSW_STEREO_3D_CTL(crtc_state->cpu_transcoder)) & S3D_ENABLE) { @@ -1123,6 +1128,7 @@ void intel_psr_short_pulse(struct intel_dp *intel_dp) if ((val & DP_PSR_SINK_STATE_MASK) == DP_PSR_SINK_INTERNAL_ERROR) { DRM_DEBUG_KMS("PSR sink internal error, disabling PSR\n"); intel_psr_disable_locked(intel_dp); + psr->sink_not_reliable = true; } if (drm_dp_dpcd_readb(&intel_dp->aux, DP_PSR_ERROR_STATUS, &val) != 1) { @@ -1140,8 +1146,10 @@ void intel_psr_short_pulse(struct intel_dp *intel_dp) if (val & ~errors) DRM_ERROR("PSR_ERROR_STATUS unhandled errors %x\n", val & ~errors); - if (val & errors) + if (val & errors) { intel_psr_disable_locked(intel_dp); + psr->sink_not_reliable = true; + } /* clear status register */ drm_dp_dpcd_writeb(&intel_dp->aux, DP_PSR_ERROR_STATUS, val); exit: From 183b8e676db2cbf80c0769e472d045bb602e4ea5 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Jos=C3=A9=20Roberto=20de=20Souza?= Date: Wed, 21 Nov 2018 14:54:39 -0800 Subject: [PATCH 06/98] drm/i915: Disable PSR when a PSR aux error happen MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit While PSR is active hardware will do aux transactions by it self to wakeup sink to receive a new frame when necessary. If that transaction is not acked by sink, hardware will trigger this interruption. So let's disable PSR as it is a hint that there is problem with this sink. The removed FIXME was asking to manually train the link but we don't need to do that as by spec sink should do a short pulse when it is out of sync with source, we just need to make sure it is awaken and the SDP header with PSR inactive set it will trigger the short pulse with a error set in the link status. v3: added workarround to fix scheduled work starvation cause by to frequent PSR error interruption v4: only setting irq_aux_error as we don't care in clear it and not using dev_priv->irq_lock as consequence. v5: rebased: using edp_psr_shift() Cc: Dhinakaran Pandiyan Reviewed-by: Rodrigo Vivi Signed-off-by: José Roberto de Souza Link: https://patchwork.freedesktop.org/patch/msgid/20181121225441.18785-4-jose.souza@intel.com --- drivers/gpu/drm/i915/i915_drv.h | 1 + drivers/gpu/drm/i915/intel_psr.c | 41 ++++++++++++++++++++++++++++---- 2 files changed, 38 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 97cbcb7b016a..f763b30f98d9 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -505,6 +505,7 @@ struct i915_psr { ktime_t last_entry_attempt; ktime_t last_exit; bool sink_not_reliable; + bool irq_aux_error; }; enum intel_pch { diff --git a/drivers/gpu/drm/i915/intel_psr.c b/drivers/gpu/drm/i915/intel_psr.c index ab527f9a5436..f0bfe0a5ff7c 100644 --- a/drivers/gpu/drm/i915/intel_psr.c +++ b/drivers/gpu/drm/i915/intel_psr.c @@ -169,6 +169,7 @@ void intel_psr_irq_handler(struct drm_i915_private *dev_priv, u32 psr_iir) u32 transcoders = BIT(TRANSCODER_EDP); enum transcoder cpu_transcoder; ktime_t time_ns = ktime_get(); + u32 mask = 0; if (INTEL_GEN(dev_priv) >= 8) transcoders |= BIT(TRANSCODER_A) | @@ -178,10 +179,22 @@ void intel_psr_irq_handler(struct drm_i915_private *dev_priv, u32 psr_iir) for_each_cpu_transcoder_masked(dev_priv, cpu_transcoder, transcoders) { int shift = edp_psr_shift(cpu_transcoder); - /* FIXME: Exit PSR and link train manually when this happens. */ - if (psr_iir & EDP_PSR_ERROR(shift)) - DRM_DEBUG_KMS("[transcoder %s] PSR aux error\n", - transcoder_name(cpu_transcoder)); + if (psr_iir & EDP_PSR_ERROR(shift)) { + DRM_WARN("[transcoder %s] PSR aux error\n", + transcoder_name(cpu_transcoder)); + + dev_priv->psr.irq_aux_error = true; + + /* + * If this interruption is not masked it will keep + * interrupting so fast that it prevents the scheduled + * work to run. + * Also after a PSR error, we don't want to arm PSR + * again so we don't care about unmask the interruption + * or unset irq_aux_error. + */ + mask |= EDP_PSR_ERROR(shift); + } if (psr_iir & EDP_PSR_PRE_ENTRY(shift)) { dev_priv->psr.last_entry_attempt = time_ns; @@ -203,6 +216,13 @@ void intel_psr_irq_handler(struct drm_i915_private *dev_priv, u32 psr_iir) } } } + + if (mask) { + mask |= I915_READ(EDP_PSR_IMR); + I915_WRITE(EDP_PSR_IMR, mask); + + schedule_work(&dev_priv->psr.work); + } } static bool intel_dp_get_colorimetry_status(struct intel_dp *intel_dp) @@ -938,6 +958,16 @@ int intel_psr_set_debugfs_mode(struct drm_i915_private *dev_priv, return ret; } +static void intel_psr_handle_irq(struct drm_i915_private *dev_priv) +{ + struct i915_psr *psr = &dev_priv->psr; + + intel_psr_disable_locked(psr->dp); + psr->sink_not_reliable = true; + /* let's make sure that sink is awaken */ + drm_dp_dpcd_writeb(&psr->dp->aux, DP_SET_POWER, DP_SET_POWER_D0); +} + static void intel_psr_work(struct work_struct *work) { struct drm_i915_private *dev_priv = @@ -948,6 +978,9 @@ static void intel_psr_work(struct work_struct *work) if (!dev_priv->psr.enabled) goto unlock; + if (READ_ONCE(dev_priv->psr.irq_aux_error)) + intel_psr_handle_irq(dev_priv); + /* * We have to make sure PSR is ready for re-enable * otherwise it keeps disabled until next full enable/disable cycle. From 888bf84dba3d60a347d67ab0b6c22f72dec7eb26 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Jos=C3=A9=20Roberto=20de=20Souza?= Date: Wed, 21 Nov 2018 14:54:40 -0800 Subject: [PATCH 07/98] drm/i915: Keep PSR disabled after a driver reload after a PSR error MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit If a PSR error happened and the driver is reloaded, the EDP_PSR_IIR will still keep the error set even after the reset done in the irq_preinstall and irq_uninstall hooks. And enabling in this situation cause the screen to freeze in the first time that PSR HW tries to activate so lets keep PSR disabled to avoid any rendering problems. v5: rebased: using edp_psr_shift() v4: Moved handling from intel_psr_compute_config() to intel_psr_init() to avoid hardware access during compute(Ville) Cc: Ville Syrjälä Cc: Dhinakaran Pandiyan Reviewed-by: Rodrigo Vivi Signed-off-by: José Roberto de Souza squash Link: https://patchwork.freedesktop.org/patch/msgid/20181121225441.18785-5-jose.souza@intel.com --- drivers/gpu/drm/i915/intel_psr.c | 18 ++++++++++++++++++ 1 file changed, 18 insertions(+) diff --git a/drivers/gpu/drm/i915/intel_psr.c b/drivers/gpu/drm/i915/intel_psr.c index f0bfe0a5ff7c..c5ab69e6c90b 100644 --- a/drivers/gpu/drm/i915/intel_psr.c +++ b/drivers/gpu/drm/i915/intel_psr.c @@ -1111,6 +1111,8 @@ void intel_psr_flush(struct drm_i915_private *dev_priv, */ void intel_psr_init(struct drm_i915_private *dev_priv) { + u32 val; + if (!HAS_PSR(dev_priv)) return; @@ -1124,6 +1126,22 @@ void intel_psr_init(struct drm_i915_private *dev_priv) if (INTEL_GEN(dev_priv) < 9 || !dev_priv->vbt.psr.enable) i915_modparams.enable_psr = 0; + /* + * If a PSR error happened and the driver is reloaded, the EDP_PSR_IIR + * will still keep the error set even after the reset done in the + * irq_preinstall and irq_uninstall hooks. + * And enabling in this situation cause the screen to freeze in the + * first time that PSR HW tries to activate so lets keep PSR disabled + * to avoid any rendering problems. + */ + val = I915_READ(EDP_PSR_IIR); + val &= EDP_PSR_ERROR(edp_psr_shift(TRANSCODER_EDP)); + if (val) { + DRM_DEBUG_KMS("PSR interruption error set\n"); + dev_priv->psr.sink_not_reliable = true; + return; + } + /* Set link_standby x link_off defaults */ if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) /* HSW and BDW require workarounds that we don't implement. */ From 16c36c4cb84de069cf6b1398b7a2de29052b8eb9 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Jos=C3=A9=20Roberto=20de=20Souza?= Date: Wed, 21 Nov 2018 14:54:41 -0800 Subject: [PATCH 08/98] drm/i915/hsw: Drop the stereo 3D enabled check in psr_compute_config() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit We should not access hardware while computing config also we don't support stereo 3D so this test was never true. Suggested-by: Ville Syrjälä Cc: Ville Syrjälä Reviewed-by: Rodrigo Vivi Signed-off-by: José Roberto de Souza Link: https://patchwork.freedesktop.org/patch/msgid/20181121225441.18785-6-jose.souza@intel.com --- drivers/gpu/drm/i915/intel_psr.c | 7 ------- 1 file changed, 7 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_psr.c b/drivers/gpu/drm/i915/intel_psr.c index c5ab69e6c90b..572e626eadff 100644 --- a/drivers/gpu/drm/i915/intel_psr.c +++ b/drivers/gpu/drm/i915/intel_psr.c @@ -552,13 +552,6 @@ void intel_psr_compute_config(struct intel_dp *intel_dp, return; } - if (IS_HASWELL(dev_priv) && - I915_READ(HSW_STEREO_3D_CTL(crtc_state->cpu_transcoder)) & - S3D_ENABLE) { - DRM_DEBUG_KMS("PSR condition failed: Stereo 3D is Enabled\n"); - return; - } - if (IS_HASWELL(dev_priv) && adjusted_mode->flags & DRM_MODE_FLAG_INTERLACE) { DRM_DEBUG_KMS("PSR condition failed: Interlaced is Enabled\n"); From 0e39037b3165567660b0e03f67534da5269a0465 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Fri, 23 Nov 2018 13:23:25 +0000 Subject: [PATCH 09/98] drm/i915: Cache the error string Currently, we convert the error state into a string every time we read from sysfs (and sysfs reads in page size (4KiB) chunks). We do try to window the string and only capture the portion that is being read, but that means that we must always convert up to the window to find the start. For a very large error state bordering on EXEC_OBJECT_CAPTURE abuse, this is noticeable as it degrades to O(N^2)! As we do not have a convenient hook for sysfs open(), and we would like to keep the lazy conversion into a string, do the conversion of the whole string on the first read and keep the string until the error state is freed. v2: Don't double advance simple_read_from_buffer v3: Due to extreme pain of lack of vrealloc, use a scatterlist v4: Keep the forward iterator loosely cached v5: Stylistic improvements to reduce patch size Reported-by: Jason Ekstrand Testcase: igt/gem_exec_capture/many* Signed-off-by: Chris Wilson Cc: Mika Kuoppala Cc: Jason Ekstrand Reviewed-by: Joonas Lahtinen Link: https://patchwork.freedesktop.org/patch/msgid/20181123132325.26541-1-chris@chris-wilson.co.uk --- drivers/gpu/drm/i915/i915_debugfs.c | 28 +-- drivers/gpu/drm/i915/i915_gpu_error.c | 343 ++++++++++++++++---------- drivers/gpu/drm/i915/i915_gpu_error.h | 28 +-- drivers/gpu/drm/i915/i915_sysfs.c | 27 +- 4 files changed, 247 insertions(+), 179 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c index c120d87df594..596810e0cfe8 100644 --- a/drivers/gpu/drm/i915/i915_debugfs.c +++ b/drivers/gpu/drm/i915/i915_debugfs.c @@ -943,30 +943,30 @@ static int i915_gem_fence_regs_info(struct seq_file *m, void *data) static ssize_t gpu_state_read(struct file *file, char __user *ubuf, size_t count, loff_t *pos) { - struct i915_gpu_state *error = file->private_data; - struct drm_i915_error_state_buf str; + struct i915_gpu_state *error; ssize_t ret; - loff_t tmp; + void *buf; + error = file->private_data; if (!error) return 0; - ret = i915_error_state_buf_init(&str, error->i915, count, *pos); - if (ret) - return ret; + /* Bounce buffer required because of kernfs __user API convenience. */ + buf = kmalloc(count, GFP_KERNEL); + if (!buf) + return -ENOMEM; - ret = i915_error_state_to_str(&str, error); - if (ret) + ret = i915_gpu_state_copy_to_buffer(error, buf, *pos, count); + if (ret <= 0) goto out; - tmp = 0; - ret = simple_read_from_buffer(ubuf, count, &tmp, str.buf, str.bytes); - if (ret < 0) - goto out; + if (!copy_to_user(ubuf, buf, ret)) + *pos += ret; + else + ret = -EFAULT; - *pos = str.start + ret; out: - i915_error_state_buf_release(&str); + kfree(buf); return ret; } diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c index 8123bf0e4807..a6885a59568b 100644 --- a/drivers/gpu/drm/i915/i915_gpu_error.c +++ b/drivers/gpu/drm/i915/i915_gpu_error.c @@ -27,11 +27,14 @@ * */ -#include -#include -#include -#include #include +#include +#include +#include +#include +#include + +#include #include "i915_gpu_error.h" #include "i915_drv.h" @@ -77,112 +80,110 @@ static const char *purgeable_flag(int purgeable) return purgeable ? " purgeable" : ""; } -static bool __i915_error_ok(struct drm_i915_error_state_buf *e) +static void __sg_set_buf(struct scatterlist *sg, + void *addr, unsigned int len, loff_t it) { - - if (!e->err && WARN(e->bytes > (e->size - 1), "overflow")) { - e->err = -ENOSPC; - return false; - } - - if (e->bytes == e->size - 1 || e->err) - return false; - - return true; + sg->page_link = (unsigned long)virt_to_page(addr); + sg->offset = offset_in_page(addr); + sg->length = len; + sg->dma_address = it; } -static bool __i915_error_seek(struct drm_i915_error_state_buf *e, - unsigned len) +static bool __i915_error_grow(struct drm_i915_error_state_buf *e, size_t len) { - if (e->pos + len <= e->start) { - e->pos += len; + if (!len) return false; + + if (e->bytes + len + 1 <= e->size) + return true; + + if (e->bytes) { + __sg_set_buf(e->cur++, e->buf, e->bytes, e->iter); + e->iter += e->bytes; + e->buf = NULL; + e->bytes = 0; } - /* First vsnprintf needs to fit in its entirety for memmove */ - if (len >= e->size) { - e->err = -EIO; - return false; - } + if (e->cur == e->end) { + struct scatterlist *sgl; - return true; -} - -static void __i915_error_advance(struct drm_i915_error_state_buf *e, - unsigned len) -{ - /* If this is first printf in this window, adjust it so that - * start position matches start of the buffer - */ - - if (e->pos < e->start) { - const size_t off = e->start - e->pos; - - /* Should not happen but be paranoid */ - if (off > len || e->bytes) { - e->err = -EIO; - return; + sgl = (typeof(sgl))__get_free_page(GFP_KERNEL); + if (!sgl) { + e->err = -ENOMEM; + return false; } - memmove(e->buf, e->buf + off, len - off); - e->bytes = len - off; - e->pos = e->start; - return; + if (e->cur) { + e->cur->offset = 0; + e->cur->length = 0; + e->cur->page_link = + (unsigned long)sgl | SG_CHAIN; + } else { + e->sgl = sgl; + } + + e->cur = sgl; + e->end = sgl + SG_MAX_SINGLE_ALLOC - 1; } - e->bytes += len; - e->pos += len; + e->size = ALIGN(len + 1, SZ_64K); + e->buf = kmalloc(e->size, GFP_KERNEL | __GFP_NOWARN | __GFP_NORETRY); + if (!e->buf) { + e->size = PAGE_ALIGN(len + 1); + e->buf = kmalloc(e->size, GFP_KERNEL); + } + if (!e->buf) { + e->err = -ENOMEM; + return false; + } + + return true; } __printf(2, 0) static void i915_error_vprintf(struct drm_i915_error_state_buf *e, - const char *f, va_list args) + const char *fmt, va_list args) { - unsigned len; + va_list ap; + int len; - if (!__i915_error_ok(e)) + if (e->err) return; - /* Seek the first printf which is hits start position */ - if (e->pos < e->start) { - va_list tmp; - - va_copy(tmp, args); - len = vsnprintf(NULL, 0, f, tmp); - va_end(tmp); - - if (!__i915_error_seek(e, len)) - return; + va_copy(ap, args); + len = vsnprintf(NULL, 0, fmt, ap); + va_end(ap); + if (len <= 0) { + e->err = len; + return; } - len = vsnprintf(e->buf + e->bytes, e->size - e->bytes, f, args); - if (len >= e->size - e->bytes) - len = e->size - e->bytes - 1; + if (!__i915_error_grow(e, len)) + return; - __i915_error_advance(e, len); + GEM_BUG_ON(e->bytes >= e->size); + len = vscnprintf(e->buf + e->bytes, e->size - e->bytes, fmt, args); + if (len < 0) { + e->err = len; + return; + } + e->bytes += len; } -static void i915_error_puts(struct drm_i915_error_state_buf *e, - const char *str) +static void i915_error_puts(struct drm_i915_error_state_buf *e, const char *str) { unsigned len; - if (!__i915_error_ok(e)) + if (e->err || !str) return; len = strlen(str); + if (!__i915_error_grow(e, len)) + return; - /* Seek the first printf which is hits start position */ - if (e->pos < e->start) { - if (!__i915_error_seek(e, len)) - return; - } - - if (len >= e->size - e->bytes) - len = e->size - e->bytes - 1; + GEM_BUG_ON(e->bytes + len > e->size); memcpy(e->buf + e->bytes, str, len); - - __i915_error_advance(e, len); + e->bytes += len; } #define err_printf(e, ...) i915_error_printf(e, __VA_ARGS__) @@ -268,6 +269,8 @@ static int compress_page(struct compress *c, if (zlib_deflate(zstream, Z_NO_FLUSH) != Z_OK) return -EIO; + + touch_nmi_watchdog(); } while (zstream->avail_in); /* Fallback to uncompressed if we increase size? */ @@ -635,22 +638,30 @@ static void err_print_uc(struct drm_i915_error_state_buf *m, print_error_obj(m, NULL, "GuC log buffer", error_uc->guc_log); } -int i915_error_state_to_str(struct drm_i915_error_state_buf *m, - const struct i915_gpu_state *error) +static void err_free_sgl(struct scatterlist *sgl) +{ + while (sgl) { + struct scatterlist *sg; + + for (sg = sgl; !sg_is_chain(sg); sg++) { + kfree(sg_virt(sg)); + if (sg_is_last(sg)) + break; + } + + sg = sg_is_last(sg) ? NULL : sg_chain_ptr(sg); + free_page((unsigned long)sgl); + sgl = sg; + } +} + +static void __err_print_to_sgl(struct drm_i915_error_state_buf *m, + struct i915_gpu_state *error) { - struct drm_i915_private *dev_priv = m->i915; struct drm_i915_error_object *obj; struct timespec64 ts; int i, j; - if (!error) { - err_printf(m, "No error state collected\n"); - return 0; - } - - if (IS_ERR(error)) - return PTR_ERR(error); - if (*error->error_msg) err_printf(m, "%s\n", error->error_msg); err_printf(m, "Kernel: %s\n", init_utsname()->release); @@ -683,12 +694,12 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m, err_printf(m, "Reset count: %u\n", error->reset_count); err_printf(m, "Suspend count: %u\n", error->suspend_count); err_printf(m, "Platform: %s\n", intel_platform_name(error->device_info.platform)); - err_print_pciid(m, error->i915); + err_print_pciid(m, m->i915); err_printf(m, "IOMMU enabled?: %d\n", error->iommu); - if (HAS_CSR(dev_priv)) { - struct intel_csr *csr = &dev_priv->csr; + if (HAS_CSR(m->i915)) { + struct intel_csr *csr = &m->i915->csr; err_printf(m, "DMC loaded: %s\n", yesno(csr->dmc_payload != NULL)); @@ -708,22 +719,23 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m, err_printf(m, "FORCEWAKE: 0x%08x\n", error->forcewake); err_printf(m, "DERRMR: 0x%08x\n", error->derrmr); err_printf(m, "CCID: 0x%08x\n", error->ccid); - err_printf(m, "Missed interrupts: 0x%08lx\n", dev_priv->gpu_error.missed_irq_rings); + err_printf(m, "Missed interrupts: 0x%08lx\n", + m->i915->gpu_error.missed_irq_rings); for (i = 0; i < error->nfence; i++) err_printf(m, " fence[%d] = %08llx\n", i, error->fence[i]); - if (INTEL_GEN(dev_priv) >= 6) { + if (INTEL_GEN(m->i915) >= 6) { err_printf(m, "ERROR: 0x%08x\n", error->error); - if (INTEL_GEN(dev_priv) >= 8) + if (INTEL_GEN(m->i915) >= 8) err_printf(m, "FAULT_TLB_DATA: 0x%08x 0x%08x\n", error->fault_data1, error->fault_data0); err_printf(m, "DONE_REG: 0x%08x\n", error->done_reg); } - if (IS_GEN7(dev_priv)) + if (IS_GEN7(m->i915)) err_printf(m, "ERR_INT: 0x%08x\n", error->err_int); for (i = 0; i < ARRAY_SIZE(error->engine); i++) { @@ -745,7 +757,7 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m, len += scnprintf(buf + len, sizeof(buf), "%s%s", first ? "" : ", ", - dev_priv->engine[j]->name); + m->i915->engine[j]->name); first = 0; } scnprintf(buf + len, sizeof(buf), ")"); @@ -763,7 +775,7 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m, obj = ee->batchbuffer; if (obj) { - err_puts(m, dev_priv->engine[i]->name); + err_puts(m, m->i915->engine[i]->name); if (ee->context.pid) err_printf(m, " (submitted by %s [%d], ctx %d [%d], score %d%s)", ee->context.comm, @@ -775,16 +787,16 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m, err_printf(m, " --- gtt_offset = 0x%08x %08x\n", upper_32_bits(obj->gtt_offset), lower_32_bits(obj->gtt_offset)); - print_error_obj(m, dev_priv->engine[i], NULL, obj); + print_error_obj(m, m->i915->engine[i], NULL, obj); } for (j = 0; j < ee->user_bo_count; j++) - print_error_obj(m, dev_priv->engine[i], + print_error_obj(m, m->i915->engine[i], "user", ee->user_bo[j]); if (ee->num_requests) { err_printf(m, "%s --- %d requests\n", - dev_priv->engine[i]->name, + m->i915->engine[i]->name, ee->num_requests); for (j = 0; j < ee->num_requests; j++) error_print_request(m, " ", @@ -794,10 +806,10 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m, if (IS_ERR(ee->waiters)) { err_printf(m, "%s --- ? waiters [unable to acquire spinlock]\n", - dev_priv->engine[i]->name); + m->i915->engine[i]->name); } else if (ee->num_waiters) { err_printf(m, "%s --- %d waiters\n", - dev_priv->engine[i]->name, + m->i915->engine[i]->name, ee->num_waiters); for (j = 0; j < ee->num_waiters; j++) { err_printf(m, " seqno 0x%08x for %s [%d]\n", @@ -807,22 +819,22 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m, } } - print_error_obj(m, dev_priv->engine[i], + print_error_obj(m, m->i915->engine[i], "ringbuffer", ee->ringbuffer); - print_error_obj(m, dev_priv->engine[i], + print_error_obj(m, m->i915->engine[i], "HW Status", ee->hws_page); - print_error_obj(m, dev_priv->engine[i], + print_error_obj(m, m->i915->engine[i], "HW context", ee->ctx); - print_error_obj(m, dev_priv->engine[i], + print_error_obj(m, m->i915->engine[i], "WA context", ee->wa_ctx); - print_error_obj(m, dev_priv->engine[i], + print_error_obj(m, m->i915->engine[i], "WA batchbuffer", ee->wa_batchbuffer); - print_error_obj(m, dev_priv->engine[i], + print_error_obj(m, m->i915->engine[i], "NULL context", ee->default_state); } @@ -835,43 +847,107 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m, err_print_capabilities(m, &error->device_info, &error->driver_caps); err_print_params(m, &error->params); err_print_uc(m, &error->uc); +} - if (m->bytes == 0 && m->err) - return m->err; +static int err_print_to_sgl(struct i915_gpu_state *error) +{ + struct drm_i915_error_state_buf m; + + if (IS_ERR(error)) + return PTR_ERR(error); + + if (READ_ONCE(error->sgl)) + return 0; + + memset(&m, 0, sizeof(m)); + m.i915 = error->i915; + + __err_print_to_sgl(&m, error); + + if (m.buf) { + __sg_set_buf(m.cur++, m.buf, m.bytes, m.iter); + m.bytes = 0; + m.buf = NULL; + } + if (m.cur) { + GEM_BUG_ON(m.end < m.cur); + sg_mark_end(m.cur - 1); + } + GEM_BUG_ON(m.sgl && !m.cur); + + if (m.err) { + err_free_sgl(m.sgl); + return m.err; + } + + if (cmpxchg(&error->sgl, NULL, m.sgl)) + err_free_sgl(m.sgl); return 0; } -int i915_error_state_buf_init(struct drm_i915_error_state_buf *ebuf, - struct drm_i915_private *i915, - size_t count, loff_t pos) +ssize_t i915_gpu_state_copy_to_buffer(struct i915_gpu_state *error, + char *buf, loff_t off, size_t rem) { - memset(ebuf, 0, sizeof(*ebuf)); - ebuf->i915 = i915; + struct scatterlist *sg; + size_t count; + loff_t pos; + int err; - /* We need to have enough room to store any i915_error_state printf - * so that we can move it to start position. - */ - ebuf->size = count + 1 > PAGE_SIZE ? count + 1 : PAGE_SIZE; - ebuf->buf = kmalloc(ebuf->size, - GFP_KERNEL | __GFP_NORETRY | __GFP_NOWARN); + if (!error || !rem) + return 0; - if (ebuf->buf == NULL) { - ebuf->size = PAGE_SIZE; - ebuf->buf = kmalloc(ebuf->size, GFP_KERNEL); - } + err = err_print_to_sgl(error); + if (err) + return err; - if (ebuf->buf == NULL) { - ebuf->size = 128; - ebuf->buf = kmalloc(ebuf->size, GFP_KERNEL); - } + sg = READ_ONCE(error->fit); + if (!sg || off < sg->dma_address) + sg = error->sgl; + if (!sg) + return 0; - if (ebuf->buf == NULL) - return -ENOMEM; + pos = sg->dma_address; + count = 0; + do { + size_t len, start; - ebuf->start = pos; + if (sg_is_chain(sg)) { + sg = sg_chain_ptr(sg); + GEM_BUG_ON(sg_is_chain(sg)); + } - return 0; + len = sg->length; + if (pos + len <= off) { + pos += len; + continue; + } + + start = sg->offset; + if (pos < off) { + GEM_BUG_ON(off - pos > len); + len -= off - pos; + start += off - pos; + pos = off; + } + + len = min(len, rem); + GEM_BUG_ON(!len || len > sg->length); + + memcpy(buf, page_address(sg_page(sg)) + start, len); + + count += len; + pos += len; + + buf += len; + rem -= len; + if (!rem) { + WRITE_ONCE(error->fit, sg); + break; + } + } while (!sg_is_last(sg++)); + + return count; } static void i915_error_object_free(struct drm_i915_error_object *obj) @@ -944,6 +1020,7 @@ void __i915_gpu_state_free(struct kref *error_ref) cleanup_params(error); cleanup_uc_state(error); + err_free_sgl(error->sgl); kfree(error); } diff --git a/drivers/gpu/drm/i915/i915_gpu_error.h b/drivers/gpu/drm/i915/i915_gpu_error.h index 3ec89a504de5..ff2652bbb0b0 100644 --- a/drivers/gpu/drm/i915/i915_gpu_error.h +++ b/drivers/gpu/drm/i915/i915_gpu_error.h @@ -192,6 +192,8 @@ struct i915_gpu_state { } *active_bo[I915_NUM_ENGINES], *pinned_bo; u32 active_bo_count[I915_NUM_ENGINES], pinned_bo_count; struct i915_address_space *active_vm[I915_NUM_ENGINES]; + + struct scatterlist *sgl, *fit; }; struct i915_gpu_error { @@ -298,29 +300,20 @@ struct i915_gpu_error { struct drm_i915_error_state_buf { struct drm_i915_private *i915; - unsigned int bytes; - unsigned int size; + struct scatterlist *sgl, *cur, *end; + + char *buf; + size_t bytes; + size_t size; + loff_t iter; + int err; - u8 *buf; - loff_t start; - loff_t pos; }; #if IS_ENABLED(CONFIG_DRM_I915_CAPTURE_ERROR) __printf(2, 3) void i915_error_printf(struct drm_i915_error_state_buf *e, const char *f, ...); -int i915_error_state_to_str(struct drm_i915_error_state_buf *estr, - const struct i915_gpu_state *gpu); -int i915_error_state_buf_init(struct drm_i915_error_state_buf *eb, - struct drm_i915_private *i915, - size_t count, loff_t pos); - -static inline void -i915_error_state_buf_release(struct drm_i915_error_state_buf *eb) -{ - kfree(eb->buf); -} struct i915_gpu_state *i915_capture_gpu_state(struct drm_i915_private *i915); void i915_capture_error_state(struct drm_i915_private *dev_priv, @@ -334,6 +327,9 @@ i915_gpu_state_get(struct i915_gpu_state *gpu) return gpu; } +ssize_t i915_gpu_state_copy_to_buffer(struct i915_gpu_state *error, + char *buf, loff_t offset, size_t count); + void __i915_gpu_state_free(struct kref *kref); static inline void i915_gpu_state_put(struct i915_gpu_state *gpu) { diff --git a/drivers/gpu/drm/i915/i915_sysfs.c b/drivers/gpu/drm/i915/i915_sysfs.c index e5e6f6bb2b05..ae63a7d0f51d 100644 --- a/drivers/gpu/drm/i915/i915_sysfs.c +++ b/drivers/gpu/drm/i915/i915_sysfs.c @@ -516,26 +516,21 @@ static ssize_t error_state_read(struct file *filp, struct kobject *kobj, { struct device *kdev = kobj_to_dev(kobj); - struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev); - struct drm_i915_error_state_buf error_str; + struct drm_i915_private *i915 = kdev_minor_to_i915(kdev); struct i915_gpu_state *gpu; ssize_t ret; - ret = i915_error_state_buf_init(&error_str, dev_priv, count, off); - if (ret) - return ret; + gpu = i915_first_error_state(i915); + if (gpu) { + ret = i915_gpu_state_copy_to_buffer(gpu, buf, off, count); + i915_gpu_state_put(gpu); + } else { + const char *str = "No error state collected\n"; + size_t len = strlen(str); - gpu = i915_first_error_state(dev_priv); - ret = i915_error_state_to_str(&error_str, gpu); - if (ret) - goto out; - - ret = count < error_str.bytes ? count : error_str.bytes; - memcpy(buf, error_str.buf, ret); - -out: - i915_gpu_state_put(gpu); - i915_error_state_buf_release(&error_str); + ret = min_t(size_t, count, len - off); + memcpy(buf, str + off, ret); + } return ret; } From b7f21899276a3e06ea3c98d0b3771f09eefc6e3d Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Mon, 26 Nov 2018 12:28:21 +0000 Subject: [PATCH 10/98] drm/i915/ringbuffer: 2-step restart We may be simply restarting too fast for the culmudgeonly gen3/gen4 as we still see missing interrupts following a reset. So let's try restarting a little slower, first wake up the ring empty and then tell it about the work it has to perform. References: https://bugs.freedesktop.org/show_bug.cgi?id=108735 Signed-off-by: Chris Wilson Reviewed-by: Joonas Lahtinen Link: https://patchwork.freedesktop.org/patch/msgid/20181126122821.4537-1-chris@chris-wilson.co.uk --- drivers/gpu/drm/i915/intel_ringbuffer.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index 87eebc13c0d8..e18a64d41843 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -546,10 +546,11 @@ static int init_ring_common(struct intel_engine_cs *engine) /* Check that the ring offsets point within the ring! */ GEM_BUG_ON(!intel_ring_offset_valid(ring, ring->head)); GEM_BUG_ON(!intel_ring_offset_valid(ring, ring->tail)); - intel_ring_update_space(ring); + + /* First wake the ring up to an empty/idle ring */ I915_WRITE_HEAD(engine, ring->head); - I915_WRITE_TAIL(engine, ring->tail); + I915_WRITE_TAIL(engine, ring->head); (void)I915_READ_TAIL(engine); I915_WRITE_CTL(engine, RING_CTL_SIZE(ring->size) | RING_VALID); @@ -574,6 +575,12 @@ static int init_ring_common(struct intel_engine_cs *engine) if (INTEL_GEN(dev_priv) > 2) I915_WRITE_MODE(engine, _MASKED_BIT_DISABLE(STOP_RING)); + /* Now awake, let it get started */ + if (ring->tail != ring->head) { + I915_WRITE_TAIL(engine, ring->tail); + (void)I915_READ_TAIL(engine); + } + /* Papering over lost _interrupts_ immediately following the restart */ intel_engine_wakeup(engine); out: From 39e84937b5b447973f2c7322ce4da35775e1bfbf Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Mon, 26 Nov 2018 09:56:10 +0000 Subject: [PATCH 11/98] drm/i915: Skip engine serialisation for no-op seqno reset If the engine's seqno is already at our target seqno (most likely it hasn't been used since the last reset), we can skip serialising the engine and leave it as is. Signed-off-by: Chris Wilson Cc: Mika Kuoppala Reviewed-by: Mika Kuoppala Link: https://patchwork.freedesktop.org/patch/msgid/20181126095610.20962-1-chris@chris-wilson.co.uk --- drivers/gpu/drm/i915/i915_request.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c index 71107540581d..ca95ab2f4cfa 100644 --- a/drivers/gpu/drm/i915/i915_request.c +++ b/drivers/gpu/drm/i915/i915_request.c @@ -136,6 +136,9 @@ static int reset_all_global_seqno(struct drm_i915_private *i915, u32 seqno) intel_engine_get_seqno(engine), seqno); + if (seqno == engine->timeline.seqno) + continue; + kthread_park(engine->breadcrumbs.signaler); if (!i915_seqno_passed(seqno, engine->timeline.seqno)) { From 4d4101c8b32186657c4693e1c33f90679193280b Mon Sep 17 00:00:00 2001 From: Manasi Navare Date: Tue, 27 Nov 2018 13:41:03 -0800 Subject: [PATCH 12/98] drm/dsc: Modify DRM helper to return complete DSC color depth capabilities DSC DPCD color depth register advertises its color depth capabilities by setting each of the bits that corresponding to a specific color depth. This patch defines those specific color depths and adds a helper to return an array of color depth capabilities. v2: * Simplify the logic (Ville) Signed-off-by: Manasi Navare Cc: Ville Syrjala Acked-by: Sean Paul (For merging through drm-intel) Reviewed-by: Ville Syrjala Reviewed-by: Anusha Srivatsa Link: https://patchwork.freedesktop.org/patch/msgid/20181127214125.17658-1-manasi.d.navare@intel.com --- drivers/gpu/drm/drm_dp_helper.c | 14 ++++++++------ include/drm/drm_dp_helper.h | 3 ++- 2 files changed, 10 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/drm_dp_helper.c b/drivers/gpu/drm/drm_dp_helper.c index 6d483487f2b4..2d6c491a0542 100644 --- a/drivers/gpu/drm/drm_dp_helper.c +++ b/drivers/gpu/drm/drm_dp_helper.c @@ -1428,17 +1428,19 @@ u8 drm_dp_dsc_sink_line_buf_depth(const u8 dsc_dpcd[DP_DSC_RECEIVER_CAP_SIZE]) } EXPORT_SYMBOL(drm_dp_dsc_sink_line_buf_depth); -u8 drm_dp_dsc_sink_max_color_depth(const u8 dsc_dpcd[DP_DSC_RECEIVER_CAP_SIZE]) +int drm_dp_dsc_sink_supported_input_bpcs(const u8 dsc_dpcd[DP_DSC_RECEIVER_CAP_SIZE], + u8 dsc_bpc[3]) { + int num_bpc = 0; u8 color_depth = dsc_dpcd[DP_DSC_DEC_COLOR_DEPTH_CAP - DP_DSC_SUPPORT]; if (color_depth & DP_DSC_12_BPC) - return 12; + dsc_bpc[num_bpc++] = 12; if (color_depth & DP_DSC_10_BPC) - return 10; + dsc_bpc[num_bpc++] = 10; if (color_depth & DP_DSC_8_BPC) - return 8; + dsc_bpc[num_bpc++] = 8; - return 0; + return num_bpc; } -EXPORT_SYMBOL(drm_dp_dsc_sink_max_color_depth); +EXPORT_SYMBOL(drm_dp_dsc_sink_supported_input_bpcs); diff --git a/include/drm/drm_dp_helper.h b/include/drm/drm_dp_helper.h index 3314e91f6eb3..5736c942c85b 100644 --- a/include/drm/drm_dp_helper.h +++ b/include/drm/drm_dp_helper.h @@ -1123,7 +1123,8 @@ drm_dp_is_branch(const u8 dpcd[DP_RECEIVER_CAP_SIZE]) u8 drm_dp_dsc_sink_max_slice_count(const u8 dsc_dpcd[DP_DSC_RECEIVER_CAP_SIZE], bool is_edp); u8 drm_dp_dsc_sink_line_buf_depth(const u8 dsc_dpcd[DP_DSC_RECEIVER_CAP_SIZE]); -u8 drm_dp_dsc_sink_max_color_depth(const u8 dsc_dpc[DP_DSC_RECEIVER_CAP_SIZE]); +int drm_dp_dsc_sink_supported_input_bpcs(const u8 dsc_dpc[DP_DSC_RECEIVER_CAP_SIZE], + u8 dsc_bpc[3]); static inline bool drm_dp_sink_supports_dsc(const u8 dsc_dpcd[DP_DSC_RECEIVER_CAP_SIZE]) From 7c247c067590b102ed2bd119bcadf4986ca10e94 Mon Sep 17 00:00:00 2001 From: Manasi Navare Date: Tue, 27 Nov 2018 13:41:04 -0800 Subject: [PATCH 13/98] drm/dsc: Define Display Stream Compression PPS infoframe This patch defines a new header file for all the DSC 1.2 structures and creates a structure for PPS infoframe which will be used to send picture parameter set secondary data packet for display stream compression. All the PPS infoframe syntax elements are taken from DSC 1.2 specification from VESA. v4: * Remove redundant blankline in doc (Ville) * use drm_dsc namespace for all structs (Ville) * Use packed struct (Ville) v3: * Add the SPDX shorthand (Chris Wilson) v2: * Do not use bitfields in the struct (Jani Nikula) Cc: Gaurav K Singh Cc: dri-devel@lists.freedesktop.org Cc: Jani Nikula Cc: Ville Syrjala Cc: Anusha Srivatsa Cc: Harry Wentland Signed-off-by: Manasi Navare Acked-by: Sean Paul (For merging through drm-intel) Reviewed-by: Harry Wentland Link: https://patchwork.freedesktop.org/patch/msgid/20181127214125.17658-2-manasi.d.navare@intel.com --- include/drm/drm_dsc.h | 342 ++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 342 insertions(+) create mode 100644 include/drm/drm_dsc.h diff --git a/include/drm/drm_dsc.h b/include/drm/drm_dsc.h new file mode 100644 index 000000000000..78db4f61d01c --- /dev/null +++ b/include/drm/drm_dsc.h @@ -0,0 +1,342 @@ +/* SPDX-License-Identifier: MIT + * Copyright (C) 2018 Intel Corp. + * + * Authors: + * Manasi Navare + */ + +#ifndef DRM_DSC_H_ +#define DRM_DSC_H_ + +#include + +/* VESA Display Stream Compression DSC 1.2 constants */ +#define DSC_NUM_BUF_RANGES 15 + +/** + * struct picture_parameter_set - Represents 128 bytes of Picture Parameter Set + * + * The VESA DSC standard defines picture parameter set (PPS) which display + * stream compression encoders must communicate to decoders. + * The PPS is encapsulated in 128 bytes (PPS 0 through PPS 127). The fields in + * this structure are as per Table 4.1 in Vesa DSC specification v1.1/v1.2. + * The PPS fields that span over more than a byte should be stored in Big Endian + * format. + */ +struct drm_dsc_picture_parameter_set { + /** + * @dsc_version: + * PPS0[3:0] - dsc_version_minor: Contains Minor version of DSC + * PPS0[7:4] - dsc_version_major: Contains major version of DSC + */ + u8 dsc_version; + /** + * @pps_identifier: + * PPS1[7:0] - Application specific identifier that can be + * used to differentiate between different PPS tables. + */ + u8 pps_identifier; + /** + * @pps_reserved: + * PPS2[7:0]- RESERVED Byte + */ + u8 pps_reserved; + /** + * @pps_3: + * PPS3[3:0] - linebuf_depth: Contains linebuffer bit depth used to + * generate the bitstream. (0x0 - 16 bits for DSC 1.2, 0x8 - 8 bits, + * 0xA - 10 bits, 0xB - 11 bits, 0xC - 12 bits, 0xD - 13 bits, + * 0xE - 14 bits for DSC1.2, 0xF - 14 bits for DSC 1.2. + * PPS3[7:4] - bits_per_component: Bits per component for the original + * pixels of the encoded picture. + * 0x0 = 16bpc (allowed only when dsc_version_minor = 0x2) + * 0x8 = 8bpc, 0xA = 10bpc, 0xC = 12bpc, 0xE = 14bpc (also + * allowed only when dsc_minor_version = 0x2) + */ + u8 pps_3; + /** + * @pps_4: + * PPS4[1:0] -These are the most significant 2 bits of + * compressed BPP bits_per_pixel[9:0] syntax element. + * PPS4[2] - vbr_enable: 0 = VBR disabled, 1 = VBR enabled + * PPS4[3] - simple_422: Indicates if decoder drops samples to + * reconstruct the 4:2:2 picture. + * PPS4[4] - Convert_rgb: Indicates if DSC color space conversion is + * active. + * PPS4[5] - blobk_pred_enable: Indicates if BP is used to code any + * groups in picture + * PPS4[7:6] - Reseved bits + */ + u8 pps_4; + /** + * @bits_per_pixel_low: + * PPS5[7:0] - This indicates the lower significant 8 bits of + * the compressed BPP bits_per_pixel[9:0] element. + */ + u8 bits_per_pixel_low; + /** + * @pic_height: + * PPS6[7:0], PPS7[7:0] -pic_height: Specifies the number of pixel rows + * within the raster. + */ + __be16 pic_height; + /** + * @pic_width: + * PPS8[7:0], PPS9[7:0] - pic_width: Number of pixel columns within + * the raster. + */ + __be16 pic_width; + /** + * @slice_height: + * PPS10[7:0], PPS11[7:0] - Slice height in units of pixels. + */ + __be16 slice_height; + /** + * @slice_width: + * PPS12[7:0], PPS13[7:0] - Slice width in terms of pixels. + */ + __be16 slice_width; + /** + * @chunk_size: + * PPS14[7:0], PPS15[7:0] - Size in units of bytes of the chunks + * that are used for slice multiplexing. + */ + __be16 chunk_size; + /** + * @initial_xmit_delay_high: + * PPS16[1:0] - Most Significant two bits of initial transmission delay. + * It specifies the number of pixel times that the encoder waits before + * transmitting data from its rate buffer. + * PPS16[7:2] - Reserved + */ + u8 initial_xmit_delay_high; + /** + * @initial_xmit_delay_low: + * PPS17[7:0] - Least significant 8 bits of initial transmission delay. + */ + u8 initial_xmit_delay_low; + /** + * @initial_dec_delay: + * + * PPS18[7:0], PPS19[7:0] - Initial decoding delay which is the number + * of pixel times that the decoder accumulates data in its rate buffer + * before starting to decode and output pixels. + */ + __be16 initial_dec_delay; + /** + * @pps20_reserved: + * + * PPS20[7:0] - Reserved + */ + u8 pps20_reserved; + /** + * @initial_scale_value: + * PPS21[5:0] - Initial rcXformScale factor used at beginning + * of a slice. + * PPS21[7:6] - Reserved + */ + u8 initial_scale_value; + /** + * @scale_increment_interval: + * PPS22[7:0], PPS23[7:0] - Number of group times between incrementing + * the rcXformScale factor at end of a slice. + */ + __be16 scale_increment_interval; + /** + * @scale_decrement_interval_high: + * PPS24[3:0] - Higher 4 bits indicating number of group times between + * decrementing the rcXformScale factor at beginning of a slice. + * PPS24[7:4] - Reserved + */ + u8 scale_decrement_interval_high; + /** + * @scale_decrement_interval_low: + * PPS25[7:0] - Lower 8 bits of scale decrement interval + */ + u8 scale_decrement_interval_low; + /** + * @pps26_reserved: + * PPS26[7:0] + */ + u8 pps26_reserved; + /** + * @first_line_bpg_offset: + * PPS27[4:0] - Number of additional bits that are allocated + * for each group on first line of a slice. + * PPS27[7:5] - Reserved + */ + u8 first_line_bpg_offset; + /** + * @nfl_bpg_offset: + * PPS28[7:0], PPS29[7:0] - Number of bits including frac bits + * deallocated for each group for groups after the first line of slice. + */ + __be16 nfl_bpg_offset; + /** + * @slice_bpg_offset: + * PPS30, PPS31[7:0] - Number of bits that are deallocated for each + * group to enforce the slice constraint. + */ + __be16 slice_bpg_offset; + /** + * @initial_offset: + * PPS32,33[7:0] - Initial value for rcXformOffset + */ + __be16 initial_offset; + /** + * @final_offset: + * PPS34,35[7:0] - Maximum end-of-slice value for rcXformOffset + */ + __be16 final_offset; + /** + * @flatness_min_qp: + * PPS36[4:0] - Minimum QP at which flatness is signaled and + * flatness QP adjustment is made. + * PPS36[7:5] - Reserved + */ + u8 flatness_min_qp; + /** + * @flatness_max_qp: + * PPS37[4:0] - Max QP at which flatness is signalled and + * the flatness adjustment is made. + * PPS37[7:5] - Reserved + */ + u8 flatness_max_qp; + /** + * @rc_model_size: + * PPS38,39[7:0] - Number of bits within RC Model. + */ + __be16 rc_model_size; + /** + * @rc_edge_factor: + * PPS40[3:0] - Ratio of current activity vs, previous + * activity to determine presence of edge. + * PPS40[7:4] - Reserved + */ + u8 rc_edge_factor; + /** + * @rc_quant_incr_limit0: + * PPS41[4:0] - QP threshold used in short term RC + * PPS41[7:5] - Reserved + */ + u8 rc_quant_incr_limit0; + /** + * @rc_quant_incr_limit1: + * PPS42[4:0] - QP threshold used in short term RC + * PPS42[7:5] - Reserved + */ + u8 rc_quant_incr_limit1; + /** + * @rc_tgt_offset: + * PPS43[3:0] - Lower end of the variability range around the target + * bits per group that is allowed by short term RC. + * PPS43[7:4]- Upper end of the variability range around the target + * bits per group that i allowed by short term rc. + */ + u8 rc_tgt_offset; + /** + * @rc_buf_thresh: + * PPS44[7:0] - PPS57[7:0] - Specifies the thresholds in RC model for + * the 15 ranges defined by 14 thresholds. + */ + u8 rc_buf_thresh[DSC_NUM_BUF_RANGES - 1]; + /** + * @rc_range_parameters: + * PPS58[7:0] - PPS87[7:0] + * Parameters that correspond to each of the 15 ranges. + */ + __be16 rc_range_parameters[DSC_NUM_BUF_RANGES]; + /** + * @native_422_420: + * PPS88[0] - 0 = Native 4:2:2 not used + * 1 = Native 4:2:2 used + * PPS88[1] - 0 = Native 4:2:0 not use + * 1 = Native 4:2:0 used + * PPS88[7:2] - Reserved 6 bits + */ + u8 native_422_420; + /** + * @second_line_bpg_offset: + * PPS89[4:0] - Additional bits/group budget for the + * second line of a slice in Native 4:2:0 mode. + * Set to 0 if DSC minor version is 1 or native420 is 0. + * PPS89[7:5] - Reserved + */ + u8 second_line_bpg_offset; + /** + * @nsl_bpg_offset: + * PPS90[7:0], PPS91[7:0] - Number of bits that are deallocated + * for each group that is not in the second line of a slice. + */ + __be16 nsl_bpg_offset; + /** + * @second_line_offset_adj: + * PPS92[7:0], PPS93[7:0] - Used as offset adjustment for the second + * line in Native 4:2:0 mode. + */ + __be16 second_line_offset_adj; + /** + * @pps_long_94_reserved: + * PPS 94, 95, 96, 97 - Reserved + */ + u32 pps_long_94_reserved; + /** + * @pps_long_98_reserved: + * PPS 98, 99, 100, 101 - Reserved + */ + u32 pps_long_98_reserved; + /** + * @pps_long_102_reserved: + * PPS 102, 103, 104, 105 - Reserved + */ + u32 pps_long_102_reserved; + /** + * @pps_long_106_reserved: + * PPS 106, 107, 108, 109 - reserved + */ + u32 pps_long_106_reserved; + /** + * @pps_long_110_reserved: + * PPS 110, 111, 112, 113 - reserved + */ + u32 pps_long_110_reserved; + /** + * @pps_long_114_reserved: + * PPS 114 - 117 - reserved + */ + u32 pps_long_114_reserved; + /** + * @pps_long_118_reserved: + * PPS 118 - 121 - reserved + */ + u32 pps_long_118_reserved; + /** + * @pps_long_122_reserved: + * PPS 122- 125 - reserved + */ + u32 pps_long_122_reserved; + /** + * @pps_short_126_reserved: + * PPS 126, 127 - reserved + */ + __be16 pps_short_126_reserved; +} __packed; + +/** + * struct drm_dsc_pps_infoframe - DSC infoframe carrying the Picture Parameter + * Set Metadata + * + * This structure represents the DSC PPS infoframe required to send the Picture + * Parameter Set metadata required before enabling VESA Display Stream + * Compression. This is based on the DP Secondary Data Packet structure and + * comprises of SDP Header as defined in drm_dp_helper.h and PPS payload. + * + * @pps_header: Header for PPS as per DP SDP header format + * @pps_payload: PPS payload fields as per DSC specification Table 4-1 + */ +struct drm_dsc_pps_infoframe { + struct dp_sdp_header pps_header; + struct drm_dsc_picture_parameter_set pps_payload; +} __packed; + +#endif /* _DRM_DSC_H_ */ From 19fd5adbb595f022c8f3a05133a673143c79fa91 Mon Sep 17 00:00:00 2001 From: Manasi Navare Date: Tue, 27 Nov 2018 13:41:05 -0800 Subject: [PATCH 14/98] drm/dsc: Define VESA Display Stream Compression Capabilities This defines all the DSC parameters as per the VESA DSC spec that will be required for DSC encoder/decoder v6: (From Manasi) * Add a bit mask for RANGE_BPG_OFFSET for 6 bits(Manasi) v5 (From Manasi) * Add the RC constants as per the spec v4 (From Manasi) * Add the DSC_MUX_WORD_SIZE constants (Manasi) v3 (From Manasi) * Remove the duplicate define (Suggested By:Harry Wentland) v2: Define this struct in DRM (From Manasi) * Changed the data types to u8/u16 instead of unsigned longs (Manasi) * Remove driver specific fields (Manasi) * Move this struct definition to DRM (Manasi) * Define DSC 1.2 parameters (Manasi) * Use DSC_NUM_BUF_RANGES (Manasi) * Call it drm_dsc_config (Manasi) Cc: dri-devel@lists.freedesktop.org Cc: Jani Nikula Cc: Ville Syrjala Cc: Anusha Srivatsa Cc: Harry Wentland Signed-off-by: Manasi Navare Signed-off-by: Gaurav K Singh Co-developed-by: Gaurav K Singh Acked-by: Harry Wentland Acked-by: Sean Paul (For merging through drm-intel) Reviewed-by: Anusha Srivatsa Link: https://patchwork.freedesktop.org/patch/msgid/20181127214125.17658-3-manasi.d.navare@intel.com --- include/drm/drm_dsc.h | 115 +++++++++++++++++++++++++++++++++++++++++- 1 file changed, 114 insertions(+), 1 deletion(-) diff --git a/include/drm/drm_dsc.h b/include/drm/drm_dsc.h index 78db4f61d01c..3292dfed9d0a 100644 --- a/include/drm/drm_dsc.h +++ b/include/drm/drm_dsc.h @@ -11,7 +11,120 @@ #include /* VESA Display Stream Compression DSC 1.2 constants */ -#define DSC_NUM_BUF_RANGES 15 +#define DSC_NUM_BUF_RANGES 15 +#define DSC_MUX_WORD_SIZE_8_10_BPC 48 +#define DSC_MUX_WORD_SIZE_12_BPC 64 +#define DSC_RC_PIXELS_PER_GROUP 3 +#define DSC_SCALE_DECREMENT_INTERVAL_MAX 4095 +#define DSC_RANGE_BPG_OFFSET_MASK 0x3f + +/* Configuration for a single Rate Control model range */ +struct drm_dsc_rc_range_parameters { + /* Min Quantization Parameters allowed for this range */ + u8 range_min_qp; + /* Max Quantization Parameters allowed for this range */ + u8 range_max_qp; + /* Bits/group offset to apply to target for this group */ + u8 range_bpg_offset; +}; + +struct drm_dsc_config { + /* Bits / component for previous reconstructed line buffer */ + u8 line_buf_depth; + /* Bits per component to code (must be 8, 10, or 12) */ + u8 bits_per_component; + /* + * Flag indicating to do RGB - YCoCg conversion + * and back (should be 1 for RGB input) + */ + bool convert_rgb; + u8 slice_count; + /* Slice Width */ + u16 slice_width; + /* Slice Height */ + u16 slice_height; + /* + * 4:2:2 enable mode (from PPS, 4:2:2 conversion happens + * outside of DSC encode/decode algorithm) + */ + bool enable422; + /* Picture Width */ + u16 pic_width; + /* Picture Height */ + u16 pic_height; + /* Offset to bits/group used by RC to determine QP adjustment */ + u8 rc_tgt_offset_high; + /* Offset to bits/group used by RC to determine QP adjustment */ + u8 rc_tgt_offset_low; + /* Bits/pixel target << 4 (ie., 4 fractional bits) */ + u16 bits_per_pixel; + /* + * Factor to determine if an edge is present based + * on the bits produced + */ + u8 rc_edge_factor; + /* Slow down incrementing once the range reaches this value */ + u8 rc_quant_incr_limit1; + /* Slow down incrementing once the range reaches this value */ + u8 rc_quant_incr_limit0; + /* Number of pixels to delay the initial transmission */ + u16 initial_xmit_delay; + /* Number of pixels to delay the VLD on the decoder,not including SSM */ + u16 initial_dec_delay; + /* Block prediction enable */ + bool block_pred_enable; + /* Bits/group offset to use for first line of the slice */ + u8 first_line_bpg_offset; + /* Value to use for RC model offset at slice start */ + u16 initial_offset; + /* Thresholds defining each of the buffer ranges */ + u16 rc_buf_thresh[DSC_NUM_BUF_RANGES - 1]; + /* Parameters for each of the RC ranges */ + struct drm_dsc_rc_range_parameters rc_range_params[DSC_NUM_BUF_RANGES]; + /* Total size of RC model */ + u16 rc_model_size; + /* Minimum QP where flatness information is sent */ + u8 flatness_min_qp; + /* Maximum QP where flatness information is sent */ + u8 flatness_max_qp; + /* Initial value for scale factor */ + u8 initial_scale_value; + /* Decrement scale factor every scale_decrement_interval groups */ + u16 scale_decrement_interval; + /* Increment scale factor every scale_increment_interval groups */ + u16 scale_increment_interval; + /* Non-first line BPG offset to use */ + u16 nfl_bpg_offset; + /* BPG offset used to enforce slice bit */ + u16 slice_bpg_offset; + /* Final RC linear transformation offset value */ + u16 final_offset; + /* Enable on-off VBR (ie., disable stuffing bits) */ + bool vbr_enable; + /* Mux word size (in bits) for SSM mode */ + u8 mux_word_size; + /* + * The (max) size in bytes of the "chunks" that are + * used in slice multiplexing + */ + u16 slice_chunk_size; + /* Rate Control buffer siz in bits */ + u16 rc_bits; + /* DSC Minor Version */ + u8 dsc_version_minor; + /* DSC Major version */ + u8 dsc_version_major; + /* Native 4:2:2 support */ + bool native_422; + /* Native 4:2:0 support */ + bool native_420; + /* Additional bits/grp for seconnd line of slice for native 4:2:0 */ + u8 second_line_bpg_offset; + /* Num of bits deallocated for each grp that is not in second line of slice */ + u16 nsl_bpg_offset; + /* Offset adj fr second line in Native 4:2:0 mode */ + u16 second_line_offset_adj; +}; /** * struct picture_parameter_set - Represents 128 bytes of Picture Parameter Set From 082a7b86013c2b8af75fcbf490c55110945673d8 Mon Sep 17 00:00:00 2001 From: "Srivatsa, Anusha" Date: Tue, 27 Nov 2018 13:41:06 -0800 Subject: [PATCH 15/98] drm/dsc: Define Rate Control values that do not change over configurations DSC has some Rate Control values that remain constant across all configurations. These are as per the DSC standard. v3: * Define them in drm_dsc.h as they are DSC constants (Manasi) v2: * Add DP_DSC_ prefix (Jani Nikula) Cc: dri-devel@lists.freedesktop.org Cc: Manasi Navare Cc: Jani Nikula Cc: Ville Syrjala Cc: Gaurav K Singh Cc: Harry Wentland Signed-off-by: Srivatsa, Anusha Signed-off-by: Manasi Navare Reviewed-by: Manasi Navare Acked-by: Sean Paul (For merging through drm-intel) Link: https://patchwork.freedesktop.org/patch/msgid/20181127214125.17658-4-manasi.d.navare@intel.com --- include/drm/drm_dsc.h | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/include/drm/drm_dsc.h b/include/drm/drm_dsc.h index 3292dfed9d0a..b88e31bd9da7 100644 --- a/include/drm/drm_dsc.h +++ b/include/drm/drm_dsc.h @@ -18,6 +18,12 @@ #define DSC_SCALE_DECREMENT_INTERVAL_MAX 4095 #define DSC_RANGE_BPG_OFFSET_MASK 0x3f +/* DSC Rate Control Constants */ +#define DSC_RC_MODEL_SIZE_CONST 8192 +#define DSC_RC_EDGE_FACTOR_CONST 6 +#define DSC_RC_TGT_OFFSET_HI_CONST 3 +#define DSC_RC_TGT_OFFSET_LO_CONST 3 + /* Configuration for a single Rate Control model range */ struct drm_dsc_rc_range_parameters { /* Min Quantization Parameters allowed for this range */ From a408c857a92be8b4f9932def15737ced02564e47 Mon Sep 17 00:00:00 2001 From: Manasi Navare Date: Tue, 27 Nov 2018 13:41:07 -0800 Subject: [PATCH 16/98] drm/dsc: Add helpers for DSC picture parameter set infoframes According to Display Stream compression spec 1.2, the picture parameter set metadata is sent from source to sink device using the DP Secondary data packet. An infoframe is formed for the PPS SDP header and PPS SDP payload bytes. This patch adds helpers to fill the PPS SDP header and PPS SDP payload according to the DSC 1.2 specification. v7: * Use BUILD_BUG_ON() to protect changing struct size (Ville) * Remove typecaseting (Ville) * Include byteorder.h in drm_dsc.c (Ville) * Correct kernel doc spacing (Anusha) v6: * Use proper sequence points for breaking down the assignments (Chris Wilson) * Use SPDX identifier v5: Do not use bitfields for DRM structs (Jani N) v4: * Use DSC constants for params that dont change across configurations v3: * Add reference to added kernel-docs in Documentation/gpu/drm-kms-helpers.rst (Daniel Vetter) v2: * Add EXPORT_SYMBOL for the drm functions (Manasi) Cc: dri-devel@lists.freedesktop.org Cc: Jani Nikula Cc: Ville Syrjala Cc: Anusha Srivatsa Cc: Harry Wentland Signed-off-by: Manasi Navare Acked-by: Harry Wentland Reviewed-by: Anusha Srivatsa Acked-by: Sean Paul (For merging through drm-intel) Link: https://patchwork.freedesktop.org/patch/msgid/20181127214125.17658-5-manasi.d.navare@intel.com --- Documentation/gpu/drm-kms-helpers.rst | 12 ++ drivers/gpu/drm/Makefile | 2 +- drivers/gpu/drm/drm_dsc.c | 228 ++++++++++++++++++++++++++ include/drm/drm_dsc.h | 21 +++ 4 files changed, 262 insertions(+), 1 deletion(-) create mode 100644 drivers/gpu/drm/drm_dsc.c diff --git a/Documentation/gpu/drm-kms-helpers.rst b/Documentation/gpu/drm-kms-helpers.rst index 4b4dc236ef6f..b422eb8edf16 100644 --- a/Documentation/gpu/drm-kms-helpers.rst +++ b/Documentation/gpu/drm-kms-helpers.rst @@ -232,6 +232,18 @@ MIPI DSI Helper Functions Reference .. kernel-doc:: drivers/gpu/drm/drm_mipi_dsi.c :export: +Display Stream Compression Helper Functions Reference +===================================================== + +.. kernel-doc:: drivers/gpu/drm/drm_dsc.c + :doc: dsc helpers + +.. kernel-doc:: include/drm/drm_dsc.h + :internal: + +.. kernel-doc:: drivers/gpu/drm/drm_dsc.c + :export: + Output Probing Helper Functions Reference ========================================= diff --git a/drivers/gpu/drm/Makefile b/drivers/gpu/drm/Makefile index 7f3be3506057..65a643da48d7 100644 --- a/drivers/gpu/drm/Makefile +++ b/drivers/gpu/drm/Makefile @@ -32,7 +32,7 @@ drm-$(CONFIG_AGP) += drm_agpsupport.o drm-$(CONFIG_DEBUG_FS) += drm_debugfs.o drm_debugfs_crc.o drm-$(CONFIG_DRM_LOAD_EDID_FIRMWARE) += drm_edid_load.o -drm_kms_helper-y := drm_crtc_helper.o drm_dp_helper.o drm_probe_helper.o \ +drm_kms_helper-y := drm_crtc_helper.o drm_dp_helper.o drm_dsc.o drm_probe_helper.o \ drm_plane_helper.o drm_dp_mst_topology.o drm_atomic_helper.o \ drm_kms_helper_common.o drm_dp_dual_mode_helper.o \ drm_simple_kms_helper.o drm_modeset_helper.o \ diff --git a/drivers/gpu/drm/drm_dsc.c b/drivers/gpu/drm/drm_dsc.c new file mode 100644 index 000000000000..bc2b23adb072 --- /dev/null +++ b/drivers/gpu/drm/drm_dsc.c @@ -0,0 +1,228 @@ +// SPDX-License-Identifier: MIT +/* + * Copyright © 2018 Intel Corp + * + * Author: + * Manasi Navare + */ + +#include +#include +#include +#include +#include +#include +#include + +/** + * DOC: dsc helpers + * + * These functions contain some common logic and helpers to deal with VESA + * Display Stream Compression standard required for DSC on Display Port/eDP or + * MIPI display interfaces. + */ + +/** + * drm_dsc_dp_pps_header_init() - Initializes the PPS Header + * for DisplayPort as per the DP 1.4 spec. + * @pps_sdp: Secondary data packet for DSC Picture Parameter Set + */ +void drm_dsc_dp_pps_header_init(struct drm_dsc_pps_infoframe *pps_sdp) +{ + memset(&pps_sdp->pps_header, 0, sizeof(pps_sdp->pps_header)); + + pps_sdp->pps_header.HB1 = DP_SDP_PPS; + pps_sdp->pps_header.HB2 = DP_SDP_PPS_HEADER_PAYLOAD_BYTES_MINUS_1; +} +EXPORT_SYMBOL(drm_dsc_dp_pps_header_init); + +/** + * drm_dsc_pps_infoframe_pack() - Populates the DSC PPS infoframe + * using the DSC configuration parameters in the order expected + * by the DSC Display Sink device. For the DSC, the sink device + * expects the PPS payload in the big endian format for the fields + * that span more than 1 byte. + * + * @pps_sdp: + * Secondary data packet for DSC Picture Parameter Set + * @dsc_cfg: + * DSC Configuration data filled by driver + */ +void drm_dsc_pps_infoframe_pack(struct drm_dsc_pps_infoframe *pps_sdp, + const struct drm_dsc_config *dsc_cfg) +{ + int i; + + /* Protect against someone accidently changing struct size */ + BUILD_BUG_ON(sizeof(pps_sdp->pps_payload) != + DP_SDP_PPS_HEADER_PAYLOAD_BYTES_MINUS_1 + 1); + + memset(&pps_sdp->pps_payload, 0, sizeof(pps_sdp->pps_payload)); + + /* PPS 0 */ + pps_sdp->pps_payload.dsc_version = + dsc_cfg->dsc_version_minor | + dsc_cfg->dsc_version_major << DSC_PPS_VERSION_MAJOR_SHIFT; + + /* PPS 1, 2 is 0 */ + + /* PPS 3 */ + pps_sdp->pps_payload.pps_3 = + dsc_cfg->line_buf_depth | + dsc_cfg->bits_per_component << DSC_PPS_BPC_SHIFT; + + /* PPS 4 */ + pps_sdp->pps_payload.pps_4 = + ((dsc_cfg->bits_per_pixel & DSC_PPS_BPP_HIGH_MASK) >> + DSC_PPS_MSB_SHIFT) | + dsc_cfg->vbr_enable << DSC_PPS_VBR_EN_SHIFT | + dsc_cfg->enable422 << DSC_PPS_SIMPLE422_SHIFT | + dsc_cfg->convert_rgb << DSC_PPS_CONVERT_RGB_SHIFT | + dsc_cfg->block_pred_enable << DSC_PPS_BLOCK_PRED_EN_SHIFT; + + /* PPS 5 */ + pps_sdp->pps_payload.bits_per_pixel_low = + (dsc_cfg->bits_per_pixel & DSC_PPS_LSB_MASK); + + /* + * The DSC panel expects the PPS packet to have big endian format + * for data spanning 2 bytes. Use a macro cpu_to_be16() to convert + * to big endian format. If format is little endian, it will swap + * bytes to convert to Big endian else keep it unchanged. + */ + + /* PPS 6, 7 */ + pps_sdp->pps_payload.pic_height = cpu_to_be16(dsc_cfg->pic_height); + + /* PPS 8, 9 */ + pps_sdp->pps_payload.pic_width = cpu_to_be16(dsc_cfg->pic_width); + + /* PPS 10, 11 */ + pps_sdp->pps_payload.slice_height = cpu_to_be16(dsc_cfg->slice_height); + + /* PPS 12, 13 */ + pps_sdp->pps_payload.slice_width = cpu_to_be16(dsc_cfg->slice_width); + + /* PPS 14, 15 */ + pps_sdp->pps_payload.chunk_size = cpu_to_be16(dsc_cfg->slice_chunk_size); + + /* PPS 16 */ + pps_sdp->pps_payload.initial_xmit_delay_high = + ((dsc_cfg->initial_xmit_delay & + DSC_PPS_INIT_XMIT_DELAY_HIGH_MASK) >> + DSC_PPS_MSB_SHIFT); + + /* PPS 17 */ + pps_sdp->pps_payload.initial_xmit_delay_low = + (dsc_cfg->initial_xmit_delay & DSC_PPS_LSB_MASK); + + /* PPS 18, 19 */ + pps_sdp->pps_payload.initial_dec_delay = + cpu_to_be16(dsc_cfg->initial_dec_delay); + + /* PPS 20 is 0 */ + + /* PPS 21 */ + pps_sdp->pps_payload.initial_scale_value = + dsc_cfg->initial_scale_value; + + /* PPS 22, 23 */ + pps_sdp->pps_payload.scale_increment_interval = + cpu_to_be16(dsc_cfg->scale_increment_interval); + + /* PPS 24 */ + pps_sdp->pps_payload.scale_decrement_interval_high = + ((dsc_cfg->scale_decrement_interval & + DSC_PPS_SCALE_DEC_INT_HIGH_MASK) >> + DSC_PPS_MSB_SHIFT); + + /* PPS 25 */ + pps_sdp->pps_payload.scale_decrement_interval_low = + (dsc_cfg->scale_decrement_interval & DSC_PPS_LSB_MASK); + + /* PPS 26[7:0], PPS 27[7:5] RESERVED */ + + /* PPS 27 */ + pps_sdp->pps_payload.first_line_bpg_offset = + dsc_cfg->first_line_bpg_offset; + + /* PPS 28, 29 */ + pps_sdp->pps_payload.nfl_bpg_offset = + cpu_to_be16(dsc_cfg->nfl_bpg_offset); + + /* PPS 30, 31 */ + pps_sdp->pps_payload.slice_bpg_offset = + cpu_to_be16(dsc_cfg->slice_bpg_offset); + + /* PPS 32, 33 */ + pps_sdp->pps_payload.initial_offset = + cpu_to_be16(dsc_cfg->initial_offset); + + /* PPS 34, 35 */ + pps_sdp->pps_payload.final_offset = cpu_to_be16(dsc_cfg->final_offset); + + /* PPS 36 */ + pps_sdp->pps_payload.flatness_min_qp = dsc_cfg->flatness_min_qp; + + /* PPS 37 */ + pps_sdp->pps_payload.flatness_max_qp = dsc_cfg->flatness_max_qp; + + /* PPS 38, 39 */ + pps_sdp->pps_payload.rc_model_size = + cpu_to_be16(DSC_RC_MODEL_SIZE_CONST); + + /* PPS 40 */ + pps_sdp->pps_payload.rc_edge_factor = DSC_RC_EDGE_FACTOR_CONST; + + /* PPS 41 */ + pps_sdp->pps_payload.rc_quant_incr_limit0 = + dsc_cfg->rc_quant_incr_limit0; + + /* PPS 42 */ + pps_sdp->pps_payload.rc_quant_incr_limit1 = + dsc_cfg->rc_quant_incr_limit1; + + /* PPS 43 */ + pps_sdp->pps_payload.rc_tgt_offset = DSC_RC_TGT_OFFSET_LO_CONST | + DSC_RC_TGT_OFFSET_HI_CONST << DSC_PPS_RC_TGT_OFFSET_HI_SHIFT; + + /* PPS 44 - 57 */ + for (i = 0; i < DSC_NUM_BUF_RANGES - 1; i++) + pps_sdp->pps_payload.rc_buf_thresh[i] = + dsc_cfg->rc_buf_thresh[i]; + + /* PPS 58 - 87 */ + /* + * For DSC sink programming the RC Range parameter fields + * are as follows: Min_qp[15:11], max_qp[10:6], offset[5:0] + */ + for (i = 0; i < DSC_NUM_BUF_RANGES; i++) { + pps_sdp->pps_payload.rc_range_parameters[i] = + ((dsc_cfg->rc_range_params[i].range_min_qp << + DSC_PPS_RC_RANGE_MINQP_SHIFT) | + (dsc_cfg->rc_range_params[i].range_max_qp << + DSC_PPS_RC_RANGE_MAXQP_SHIFT) | + (dsc_cfg->rc_range_params[i].range_bpg_offset)); + pps_sdp->pps_payload.rc_range_parameters[i] = + cpu_to_be16(pps_sdp->pps_payload.rc_range_parameters[i]); + } + + /* PPS 88 */ + pps_sdp->pps_payload.native_422_420 = dsc_cfg->native_422 | + dsc_cfg->native_420 << DSC_PPS_NATIVE_420_SHIFT; + + /* PPS 89 */ + pps_sdp->pps_payload.second_line_bpg_offset = + dsc_cfg->second_line_bpg_offset; + + /* PPS 90, 91 */ + pps_sdp->pps_payload.nsl_bpg_offset = + cpu_to_be16(dsc_cfg->nsl_bpg_offset); + + /* PPS 92, 93 */ + pps_sdp->pps_payload.second_line_offset_adj = + cpu_to_be16(dsc_cfg->second_line_offset_adj); + + /* PPS 94 - 127 are O */ +} +EXPORT_SYMBOL(drm_dsc_pps_infoframe_pack); diff --git a/include/drm/drm_dsc.h b/include/drm/drm_dsc.h index b88e31bd9da7..52e57ceaff80 100644 --- a/include/drm/drm_dsc.h +++ b/include/drm/drm_dsc.h @@ -24,6 +24,23 @@ #define DSC_RC_TGT_OFFSET_HI_CONST 3 #define DSC_RC_TGT_OFFSET_LO_CONST 3 +/* DSC PPS constants and macros */ +#define DSC_PPS_VERSION_MAJOR_SHIFT 4 +#define DSC_PPS_BPC_SHIFT 4 +#define DSC_PPS_MSB_SHIFT 8 +#define DSC_PPS_LSB_MASK (0xFF << 0) +#define DSC_PPS_BPP_HIGH_MASK (0x3 << 8) +#define DSC_PPS_VBR_EN_SHIFT 2 +#define DSC_PPS_SIMPLE422_SHIFT 3 +#define DSC_PPS_CONVERT_RGB_SHIFT 4 +#define DSC_PPS_BLOCK_PRED_EN_SHIFT 5 +#define DSC_PPS_INIT_XMIT_DELAY_HIGH_MASK (0x3 << 8) +#define DSC_PPS_SCALE_DEC_INT_HIGH_MASK (0xF << 8) +#define DSC_PPS_RC_TGT_OFFSET_HI_SHIFT 4 +#define DSC_PPS_RC_RANGE_MINQP_SHIFT 11 +#define DSC_PPS_RC_RANGE_MAXQP_SHIFT 6 +#define DSC_PPS_NATIVE_420_SHIFT 1 + /* Configuration for a single Rate Control model range */ struct drm_dsc_rc_range_parameters { /* Min Quantization Parameters allowed for this range */ @@ -458,4 +475,8 @@ struct drm_dsc_pps_infoframe { struct drm_dsc_picture_parameter_set pps_payload; } __packed; +void drm_dsc_dp_pps_header_init(struct drm_dsc_pps_infoframe *pps_sdp); +void drm_dsc_pps_infoframe_pack(struct drm_dsc_pps_infoframe *pps_sdp, + const struct drm_dsc_config *dsc_cfg); + #endif /* _DRM_DSC_H_ */ From f25310c7360b1825ee25858b68685c94abaa11a3 Mon Sep 17 00:00:00 2001 From: Manasi Navare Date: Tue, 27 Nov 2018 13:41:08 -0800 Subject: [PATCH 17/98] drm/dsc: Define the DSC 1.1 and 1.2 Line Buffer depth constants MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit DSC specification defines linebuf_depth which contains the line buffer bit depth used to generate the bitstream. These values are defined as per Table 4.1 in DSC 1.2 spec v2 (From Manasi): * Rename as MAX_LINEBUF_DEPTH for DSC 1.1 and DSC 1.2 Cc: dri-devel@lists.freedesktop.org Cc: Jani Nikula Cc: Ville Syrjälä Signed-off-by: Gaurav K Singh Signed-off-by: Manasi Navare Acked-by: Sean Paul (For merging through drm-intel) Reviewed-by: Anusha Srivatsa Link: https://patchwork.freedesktop.org/patch/msgid/20181127214125.17658-6-manasi.d.navare@intel.com --- include/drm/drm_dsc.h | 3 +++ 1 file changed, 3 insertions(+) diff --git a/include/drm/drm_dsc.h b/include/drm/drm_dsc.h index 52e57ceaff80..d03f1b83421a 100644 --- a/include/drm/drm_dsc.h +++ b/include/drm/drm_dsc.h @@ -40,6 +40,9 @@ #define DSC_PPS_RC_RANGE_MINQP_SHIFT 11 #define DSC_PPS_RC_RANGE_MAXQP_SHIFT 6 #define DSC_PPS_NATIVE_420_SHIFT 1 +#define DSC_1_2_MAX_LINEBUF_DEPTH_BITS 16 +#define DSC_1_2_MAX_LINEBUF_DEPTH_VAL 0 +#define DSC_1_1_MAX_LINEBUF_DEPTH_BITS 13 /* Configuration for a single Rate Control model range */ struct drm_dsc_rc_range_parameters { From 1aca96cc545536fc00b533764135ced79a9a65ba Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Wed, 28 Nov 2018 13:53:25 +0000 Subject: [PATCH 18/98] drm/i915: Mark up early pre-production Kabylakes Mark A0 as the one and only pre-production variant of Kabylake and remove its couple of workarounds, consigning them to the annals of history. Signed-off-by: Chris Wilson Cc: Joonas Lahtinen Cc: Jani Nikula Cc: Rodrigo Vivi Cc: Mika Kuoppala Reviewed-by: Mika Kuoppala Acked-by: Rodrigo Vivi Link: https://patchwork.freedesktop.org/patch/msgid/20181128135325.10641-1-chris@chris-wilson.co.uk --- drivers/gpu/drm/i915/i915_drv.c | 1 + drivers/gpu/drm/i915/intel_lrc.c | 12 ------------ drivers/gpu/drm/i915/intel_workarounds.c | 5 ----- 3 files changed, 1 insertion(+), 17 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c index b1d23c73c147..e39016713464 100644 --- a/drivers/gpu/drm/i915/i915_drv.c +++ b/drivers/gpu/drm/i915/i915_drv.c @@ -868,6 +868,7 @@ static void intel_detect_preproduction_hw(struct drm_i915_private *dev_priv) pre |= IS_HSW_EARLY_SDV(dev_priv); pre |= IS_SKL_REVID(dev_priv, 0, SKL_REVID_F0); pre |= IS_BXT_REVID(dev_priv, 0, BXT_REVID_B_LAST); + pre |= IS_KBL_REVID(dev_priv, 0, KBL_REVID_A0); if (pre) { DRM_ERROR("This is a pre-production stepping. " diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c index 08fd9b12e4d7..11f4e6148557 100644 --- a/drivers/gpu/drm/i915/intel_lrc.c +++ b/drivers/gpu/drm/i915/intel_lrc.c @@ -1401,18 +1401,6 @@ static u32 *gen9_init_indirectctx_bb(struct intel_engine_cs *engine, u32 *batch) batch = emit_lri(batch, lri, ARRAY_SIZE(lri)); - /* WaClearSlmSpaceAtContextSwitch:kbl */ - /* Actual scratch location is at 128 bytes offset */ - if (IS_KBL_REVID(engine->i915, 0, KBL_REVID_A0)) { - batch = gen8_emit_pipe_control(batch, - PIPE_CONTROL_FLUSH_L3 | - PIPE_CONTROL_GLOBAL_GTT_IVB | - PIPE_CONTROL_CS_STALL | - PIPE_CONTROL_QW_WRITE, - i915_ggtt_offset(engine->scratch) - + 2 * CACHELINE_BYTES); - } - /* WaMediaPoolStateCmdInWABB:bxt,glk */ if (HAS_POOLED_EU(engine->i915)) { /* diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c index ca1f78a42b17..39cd1f823ea9 100644 --- a/drivers/gpu/drm/i915/intel_workarounds.c +++ b/drivers/gpu/drm/i915/intel_workarounds.c @@ -384,11 +384,6 @@ static int kbl_ctx_workarounds_init(struct drm_i915_private *dev_priv) if (ret) return ret; - /* WaDisableFenceDestinationToSLM:kbl (pre-prod) */ - if (IS_KBL_REVID(dev_priv, KBL_REVID_A0, KBL_REVID_A0)) - WA_SET_BIT_MASKED(HDC_CHICKEN0, - HDC_FENCE_DEST_SLM_DISABLE); - /* WaToEnableHwFixForPushConstHWBug:kbl */ if (IS_KBL_REVID(dev_priv, KBL_REVID_C0, REVID_FOREVER)) WA_SET_BIT_MASKED(COMMON_SLICE_CHICKEN2, From 83234d13f9fda65e6c1a5e5900aa247334b1a621 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Wed, 14 Nov 2018 23:07:17 +0200 Subject: [PATCH 19/98] drm/i915: Reorganize plane register writes to make them more atomic MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Some observations about the plane registers: - the control register will self-arm if the plane is not already enabled, thus we want to write it as close to (or ideally after) the surface register - tileoff/linoff/offset/aux_offset are self-arming as well so we want them close to the surface register as well - color keying registers we maybe self arming before SKL. Not 100% sure but we can try to keep them near to the surface register as well - chv pipe b csc register are double buffered but self arming so moving them down a bit - the rest should be mostly armed by the surface register so we can safely write them first, and to just for some consistency let's try to follow keep them in order based on the register offset None of this will have any effect of course unless the vblank evasion fails (which it still does sometimes). Another potential future benefit might be pulling the non-self armings registers outside the vblank evasion since they won't latch until the arming register has been written. This would make the critical section a bit lighter and thus less likely to exceed the deadline. v2: Rebase due to input CSC v3: Swap LINOFF/TILEOFF and KEYMSK/KEYMAX to actually follow the last rule above (Matt) Add a bit more rationale to the commit message (Matt) Cc: Matt Roper Signed-off-by: Ville Syrjälä Link: https://patchwork.freedesktop.org/patch/msgid/20181114210729.16185-2-ville.syrjala@linux.intel.com Reviewed-by: Matt Roper Reviewed-by: Maarten Lankhorst --- drivers/gpu/drm/i915/intel_display.c | 40 +++++---- drivers/gpu/drm/i915/intel_sprite.c | 118 ++++++++++++++++----------- 2 files changed, 91 insertions(+), 67 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index e9f4e22b2a4e..3516b3d04787 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -3315,7 +3315,6 @@ static void i9xx_update_plane(struct intel_plane *plane, enum i9xx_plane_id i9xx_plane = plane->i9xx_plane; u32 linear_offset; u32 dspcntr = plane_state->ctl; - i915_reg_t reg = DSPCNTR(i9xx_plane); int x = plane_state->color_plane[0].x; int y = plane_state->color_plane[0].y; unsigned long irqflags; @@ -3330,41 +3329,45 @@ static void i9xx_update_plane(struct intel_plane *plane, spin_lock_irqsave(&dev_priv->uncore.lock, irqflags); + I915_WRITE_FW(DSPSTRIDE(i9xx_plane), plane_state->color_plane[0].stride); + if (INTEL_GEN(dev_priv) < 4) { /* pipesrc and dspsize control the size that is scaled from, * which should always be the user's requested size. */ + I915_WRITE_FW(DSPPOS(i9xx_plane), 0); I915_WRITE_FW(DSPSIZE(i9xx_plane), ((crtc_state->pipe_src_h - 1) << 16) | (crtc_state->pipe_src_w - 1)); - I915_WRITE_FW(DSPPOS(i9xx_plane), 0); } else if (IS_CHERRYVIEW(dev_priv) && i9xx_plane == PLANE_B) { + I915_WRITE_FW(PRIMPOS(i9xx_plane), 0); I915_WRITE_FW(PRIMSIZE(i9xx_plane), ((crtc_state->pipe_src_h - 1) << 16) | (crtc_state->pipe_src_w - 1)); - I915_WRITE_FW(PRIMPOS(i9xx_plane), 0); I915_WRITE_FW(PRIMCNSTALPHA(i9xx_plane), 0); } - I915_WRITE_FW(reg, dspcntr); - - I915_WRITE_FW(DSPSTRIDE(i9xx_plane), plane_state->color_plane[0].stride); if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) { - I915_WRITE_FW(DSPSURF(i9xx_plane), - intel_plane_ggtt_offset(plane_state) + - dspaddr_offset); I915_WRITE_FW(DSPOFFSET(i9xx_plane), (y << 16) | x); } else if (INTEL_GEN(dev_priv) >= 4) { + I915_WRITE_FW(DSPLINOFF(i9xx_plane), linear_offset); + I915_WRITE_FW(DSPTILEOFF(i9xx_plane), (y << 16) | x); + } + + /* + * The control register self-arms if the plane was previously + * disabled. Try to make the plane enable atomic by writing + * the control register just before the surface register. + */ + I915_WRITE_FW(DSPCNTR(i9xx_plane), dspcntr); + if (INTEL_GEN(dev_priv) >= 4) I915_WRITE_FW(DSPSURF(i9xx_plane), intel_plane_ggtt_offset(plane_state) + dspaddr_offset); - I915_WRITE_FW(DSPTILEOFF(i9xx_plane), (y << 16) | x); - I915_WRITE_FW(DSPLINOFF(i9xx_plane), linear_offset); - } else { + else I915_WRITE_FW(DSPADDR(i9xx_plane), intel_plane_ggtt_offset(plane_state) + dspaddr_offset); - } spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags); } @@ -10087,8 +10090,8 @@ static void i9xx_update_cursor(struct intel_plane *plane, * On some platforms writing CURCNTR first will also * cause CURPOS to be armed by the CURBASE write. * Without the CURCNTR write the CURPOS write would - * arm itself. Thus we always start the full update - * with a CURCNTR write. + * arm itself. Thus we always update CURCNTR before + * CURPOS. * * On other platforms CURPOS always requires the * CURBASE write to arm the update. Additonally @@ -10098,15 +10101,16 @@ static void i9xx_update_cursor(struct intel_plane *plane, * cursor that doesn't appear to move, or even change * shape. Thus we always write CURBASE. * - * CURCNTR and CUR_FBC_CTL are always - * armed by the CURBASE write only. + * The other registers are armed by by the CURBASE write + * except when the plane is getting enabled at which time + * the CURCNTR write arms the update. */ if (plane->cursor.base != base || plane->cursor.size != fbc_ctl || plane->cursor.cntl != cntl) { - I915_WRITE_FW(CURCNTR(pipe), cntl); if (HAS_CUR_FBC(dev_priv)) I915_WRITE_FW(CUR_FBC_CTL(pipe), fbc_ctl); + I915_WRITE_FW(CURCNTR(pipe), cntl); I915_WRITE_FW(CURPOS(pipe), pos); I915_WRITE_FW(CURBASE(pipe), base); diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c index abe193815ccc..6cefe5f7a374 100644 --- a/drivers/gpu/drm/i915/intel_sprite.c +++ b/drivers/gpu/drm/i915/intel_sprite.c @@ -508,28 +508,12 @@ skl_program_plane(struct intel_plane *plane, spin_lock_irqsave(&dev_priv->uncore.lock, irqflags); - if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv)) - I915_WRITE_FW(PLANE_COLOR_CTL(pipe, plane_id), - plane_state->color_ctl); - - if (fb->format->is_yuv && icl_is_hdr_plane(plane)) - icl_program_input_csc_coeff(crtc_state, plane_state); - - I915_WRITE_FW(PLANE_KEYVAL(pipe, plane_id), key->min_value); - I915_WRITE_FW(PLANE_KEYMAX(pipe, plane_id), keymax); - I915_WRITE_FW(PLANE_KEYMSK(pipe, plane_id), keymsk); - - I915_WRITE_FW(PLANE_OFFSET(pipe, plane_id), (y << 16) | x); I915_WRITE_FW(PLANE_STRIDE(pipe, plane_id), stride); + I915_WRITE_FW(PLANE_POS(pipe, plane_id), (crtc_y << 16) | crtc_x); I915_WRITE_FW(PLANE_SIZE(pipe, plane_id), (src_h << 16) | src_w); I915_WRITE_FW(PLANE_AUX_DIST(pipe, plane_id), (plane_state->color_plane[1].offset - surf_addr) | aux_stride); - if (INTEL_GEN(dev_priv) < 11) - I915_WRITE_FW(PLANE_AUX_OFFSET(pipe, plane_id), - (plane_state->color_plane[1].y << 16) | - plane_state->color_plane[1].x); - if (icl_is_hdr_plane(plane)) { u32 cus_ctl = 0; @@ -551,15 +535,36 @@ skl_program_plane(struct intel_plane *plane, I915_WRITE_FW(PLANE_CUS_CTL(pipe, plane_id), cus_ctl); } - if (!slave && plane_state->scaler_id >= 0) - skl_program_scaler(plane, crtc_state, plane_state); + if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv)) + I915_WRITE_FW(PLANE_COLOR_CTL(pipe, plane_id), + plane_state->color_ctl); - I915_WRITE_FW(PLANE_POS(pipe, plane_id), (crtc_y << 16) | crtc_x); + if (fb->format->is_yuv && icl_is_hdr_plane(plane)) + icl_program_input_csc_coeff(crtc_state, plane_state); + I915_WRITE_FW(PLANE_KEYVAL(pipe, plane_id), key->min_value); + I915_WRITE_FW(PLANE_KEYMSK(pipe, plane_id), keymsk); + I915_WRITE_FW(PLANE_KEYMAX(pipe, plane_id), keymax); + + I915_WRITE_FW(PLANE_OFFSET(pipe, plane_id), (y << 16) | x); + + if (INTEL_GEN(dev_priv) < 11) + I915_WRITE_FW(PLANE_AUX_OFFSET(pipe, plane_id), + (plane_state->color_plane[1].y << 16) | + plane_state->color_plane[1].x); + + /* + * The control register self-arms if the plane was previously + * disabled. Try to make the plane enable atomic by writing + * the control register just before the surface register. + */ I915_WRITE_FW(PLANE_CTL(pipe, plane_id), plane_ctl); I915_WRITE_FW(PLANE_SURF(pipe, plane_id), intel_plane_ggtt_offset(plane_state) + surf_addr); + if (!slave && plane_state->scaler_id >= 0) + skl_program_scaler(plane, crtc_state, plane_state); + spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags); } @@ -821,24 +826,29 @@ vlv_update_plane(struct intel_plane *plane, vlv_update_clrc(plane_state); + I915_WRITE_FW(SPSTRIDE(pipe, plane_id), + plane_state->color_plane[0].stride); + I915_WRITE_FW(SPPOS(pipe, plane_id), (crtc_y << 16) | crtc_x); + I915_WRITE_FW(SPSIZE(pipe, plane_id), (crtc_h << 16) | crtc_w); + I915_WRITE_FW(SPCONSTALPHA(pipe, plane_id), 0); + if (IS_CHERRYVIEW(dev_priv) && pipe == PIPE_B) chv_update_csc(plane_state); if (key->flags) { I915_WRITE_FW(SPKEYMINVAL(pipe, plane_id), key->min_value); - I915_WRITE_FW(SPKEYMAXVAL(pipe, plane_id), key->max_value); I915_WRITE_FW(SPKEYMSK(pipe, plane_id), key->channel_mask); + I915_WRITE_FW(SPKEYMAXVAL(pipe, plane_id), key->max_value); } - I915_WRITE_FW(SPSTRIDE(pipe, plane_id), - plane_state->color_plane[0].stride); - I915_WRITE_FW(SPPOS(pipe, plane_id), (crtc_y << 16) | crtc_x); - I915_WRITE_FW(SPTILEOFF(pipe, plane_id), (y << 16) | x); I915_WRITE_FW(SPLINOFF(pipe, plane_id), linear_offset); + I915_WRITE_FW(SPTILEOFF(pipe, plane_id), (y << 16) | x); - I915_WRITE_FW(SPCONSTALPHA(pipe, plane_id), 0); - - I915_WRITE_FW(SPSIZE(pipe, plane_id), (crtc_h << 16) | crtc_w); + /* + * The control register self-arms if the plane was previously + * disabled. Try to make the plane enable atomic by writing + * the control register just before the surface register. + */ I915_WRITE_FW(SPCNTR(pipe, plane_id), sprctl); I915_WRITE_FW(SPSURF(pipe, plane_id), intel_plane_ggtt_offset(plane_state) + sprsurf_offset); @@ -980,27 +990,32 @@ ivb_update_plane(struct intel_plane *plane, spin_lock_irqsave(&dev_priv->uncore.lock, irqflags); - if (key->flags) { - I915_WRITE_FW(SPRKEYVAL(pipe), key->min_value); - I915_WRITE_FW(SPRKEYMAX(pipe), key->max_value); - I915_WRITE_FW(SPRKEYMSK(pipe), key->channel_mask); - } - I915_WRITE_FW(SPRSTRIDE(pipe), plane_state->color_plane[0].stride); I915_WRITE_FW(SPRPOS(pipe), (crtc_y << 16) | crtc_x); + I915_WRITE_FW(SPRSIZE(pipe), (crtc_h << 16) | crtc_w); + if (IS_IVYBRIDGE(dev_priv)) + I915_WRITE_FW(SPRSCALE(pipe), sprscale); + + if (key->flags) { + I915_WRITE_FW(SPRKEYVAL(pipe), key->min_value); + I915_WRITE_FW(SPRKEYMSK(pipe), key->channel_mask); + I915_WRITE_FW(SPRKEYMAX(pipe), key->max_value); + } /* HSW consolidates SPRTILEOFF and SPRLINOFF into a single SPROFFSET * register */ if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) { I915_WRITE_FW(SPROFFSET(pipe), (y << 16) | x); } else { - I915_WRITE_FW(SPRTILEOFF(pipe), (y << 16) | x); I915_WRITE_FW(SPRLINOFF(pipe), linear_offset); + I915_WRITE_FW(SPRTILEOFF(pipe), (y << 16) | x); } - I915_WRITE_FW(SPRSIZE(pipe), (crtc_h << 16) | crtc_w); - if (IS_IVYBRIDGE(dev_priv)) - I915_WRITE_FW(SPRSCALE(pipe), sprscale); + /* + * The control register self-arms if the plane was previously + * disabled. Try to make the plane enable atomic by writing + * the control register just before the surface register. + */ I915_WRITE_FW(SPRCTL(pipe), sprctl); I915_WRITE_FW(SPRSURF(pipe), intel_plane_ggtt_offset(plane_state) + sprsurf_offset); @@ -1018,7 +1033,7 @@ ivb_disable_plane(struct intel_plane *plane, struct intel_crtc *crtc) spin_lock_irqsave(&dev_priv->uncore.lock, irqflags); I915_WRITE_FW(SPRCTL(pipe), 0); - /* Can't leave the scaler enabled... */ + /* Disable the scaler */ if (IS_IVYBRIDGE(dev_priv)) I915_WRITE_FW(SPRSCALE(pipe), 0); I915_WRITE_FW(SPRSURF(pipe), 0); @@ -1148,20 +1163,25 @@ g4x_update_plane(struct intel_plane *plane, spin_lock_irqsave(&dev_priv->uncore.lock, irqflags); - if (key->flags) { - I915_WRITE_FW(DVSKEYVAL(pipe), key->min_value); - I915_WRITE_FW(DVSKEYMAX(pipe), key->max_value); - I915_WRITE_FW(DVSKEYMSK(pipe), key->channel_mask); - } - I915_WRITE_FW(DVSSTRIDE(pipe), plane_state->color_plane[0].stride); I915_WRITE_FW(DVSPOS(pipe), (crtc_y << 16) | crtc_x); - - I915_WRITE_FW(DVSTILEOFF(pipe), (y << 16) | x); - I915_WRITE_FW(DVSLINOFF(pipe), linear_offset); - I915_WRITE_FW(DVSSIZE(pipe), (crtc_h << 16) | crtc_w); I915_WRITE_FW(DVSSCALE(pipe), dvsscale); + + if (key->flags) { + I915_WRITE_FW(DVSKEYVAL(pipe), key->min_value); + I915_WRITE_FW(DVSKEYMSK(pipe), key->channel_mask); + I915_WRITE_FW(DVSKEYMAX(pipe), key->max_value); + } + + I915_WRITE_FW(DVSLINOFF(pipe), linear_offset); + I915_WRITE_FW(DVSTILEOFF(pipe), (y << 16) | x); + + /* + * The control register self-arms if the plane was previously + * disabled. Try to make the plane enable atomic by writing + * the control register just before the surface register. + */ I915_WRITE_FW(DVSCNTR(pipe), dvscntr); I915_WRITE_FW(DVSSURF(pipe), intel_plane_ggtt_offset(plane_state) + dvssurf_offset); From 019575a58c84b1cd534d1dd1cad36592995f8f6b Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Wed, 14 Nov 2018 23:07:18 +0200 Subject: [PATCH 20/98] drm/i915: Move single buffered plane register writes to the end MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The plane color correction registers are single buffered. So ideally we would write them at the start of vblank just after the double buffered plane registers have been latched. Since we have no convenient way to do that for now let's at least move the single buffered register writes to happen after the double buffered registers have been written. Reviewed-by: Rodrigo Vivi Signed-off-by: Ville Syrjälä Link: https://patchwork.freedesktop.org/patch/msgid/20181114210729.16185-3-ville.syrjala@linux.intel.com Reviewed-by: Matt Roper Reviewed-by: Maarten Lankhorst --- drivers/gpu/drm/i915/intel_sprite.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c index 6cefe5f7a374..0548b996693f 100644 --- a/drivers/gpu/drm/i915/intel_sprite.c +++ b/drivers/gpu/drm/i915/intel_sprite.c @@ -824,8 +824,6 @@ vlv_update_plane(struct intel_plane *plane, spin_lock_irqsave(&dev_priv->uncore.lock, irqflags); - vlv_update_clrc(plane_state); - I915_WRITE_FW(SPSTRIDE(pipe, plane_id), plane_state->color_plane[0].stride); I915_WRITE_FW(SPPOS(pipe, plane_id), (crtc_y << 16) | crtc_x); @@ -853,6 +851,8 @@ vlv_update_plane(struct intel_plane *plane, I915_WRITE_FW(SPSURF(pipe, plane_id), intel_plane_ggtt_offset(plane_state) + sprsurf_offset); + vlv_update_clrc(plane_state); + spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags); } From afbd8a722bd64e0020a3b2f8ac2c42e02dc8f596 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Tue, 27 Nov 2018 18:37:42 +0200 Subject: [PATCH 21/98] drm/i915: Introduce crtc_state->update_planes bitmask MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Keep track which planes need updating during the commit. For now we set the bit for any plane that was or will be visible (including icl+ nv12 slave planes). In the future I'll have need to update invisible planes as well, for skl plane ddbs and for pre-skl pipe gamma/csc control (which lives in the primary plane control register). v2: Pimp the commit message to mention icl+ nv12 slave planes (Matt) Reviewed-by: Rodrigo Vivi Reviewed-by: Maarten Lankhorst Reviewed-by: Matt Roper Signed-off-by: Ville Syrjälä Link: https://patchwork.freedesktop.org/patch/msgid/20181127163742.30215-1-ville.syrjala@linux.intel.com --- drivers/gpu/drm/i915/intel_atomic.c | 1 + drivers/gpu/drm/i915/intel_atomic_plane.c | 8 ++++---- drivers/gpu/drm/i915/intel_display.c | 5 ++++- drivers/gpu/drm/i915/intel_drv.h | 3 +++ 4 files changed, 12 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_atomic.c b/drivers/gpu/drm/i915/intel_atomic.c index a5a2c8fe58a7..8cb02f28d30c 100644 --- a/drivers/gpu/drm/i915/intel_atomic.c +++ b/drivers/gpu/drm/i915/intel_atomic.c @@ -184,6 +184,7 @@ intel_crtc_duplicate_state(struct drm_crtc *crtc) crtc_state->fifo_changed = false; crtc_state->wm.need_postvbl_update = false; crtc_state->fb_bits = 0; + crtc_state->update_planes = 0; return &crtc_state->base; } diff --git a/drivers/gpu/drm/i915/intel_atomic_plane.c b/drivers/gpu/drm/i915/intel_atomic_plane.c index 905f8ef3ba4f..026a11511941 100644 --- a/drivers/gpu/drm/i915/intel_atomic_plane.c +++ b/drivers/gpu/drm/i915/intel_atomic_plane.c @@ -139,6 +139,9 @@ int intel_plane_atomic_check_with_state(const struct intel_crtc_state *old_crtc_ if (state->visible && state->fb->format->format == DRM_FORMAT_NV12) crtc_state->nv12_planes |= BIT(intel_plane->id); + if (state->visible || old_plane_state->base.visible) + crtc_state->update_planes |= BIT(intel_plane->id); + return intel_plane_atomic_calc_changes(old_crtc_state, &crtc_state->base, old_plane_state, @@ -173,14 +176,11 @@ void intel_update_planes_on_crtc(struct intel_atomic_state *old_state, struct intel_crtc_state *old_crtc_state, struct intel_crtc_state *new_crtc_state) { + u32 update_mask = new_crtc_state->update_planes; struct intel_plane_state *new_plane_state; struct intel_plane *plane; - u32 update_mask; int i; - update_mask = old_crtc_state->active_planes; - update_mask |= new_crtc_state->active_planes; - for_each_new_intel_plane_in_state(old_state, plane, new_plane_state, i) { if (crtc->pipe != plane->pipe || !(update_mask & BIT(plane->id))) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 3516b3d04787..4469b4511c92 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -10839,8 +10839,10 @@ static int icl_check_nv12_planes(struct intel_crtc_state *crtc_state) continue; plane_state->linked_plane = NULL; - if (plane_state->slave && !plane_state->base.visible) + if (plane_state->slave && !plane_state->base.visible) { crtc_state->active_planes &= ~BIT(plane->id); + crtc_state->update_planes |= BIT(plane->id); + } plane_state->slave = false; } @@ -10881,6 +10883,7 @@ static int icl_check_nv12_planes(struct intel_crtc_state *crtc_state) linked_state->slave = true; linked_state->linked_plane = plane; crtc_state->active_planes |= BIT(linked->id); + crtc_state->update_planes |= BIT(linked->id); DRM_DEBUG_KMS("Using %s as Y plane for %s\n", linked->base.name, plane->base.name); } diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index a62d77b76291..1f29061cefa1 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -926,6 +926,9 @@ struct intel_crtc_state { u8 active_planes; u8 nv12_planes; + /* bitmask of planes that will be updated during the commit */ + u8 update_planes; + /* HDMI scrambling status */ bool hdmi_scrambling; From 0dd14be30d4c8401be9cd129df3b3ee238ea6f78 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Wed, 14 Nov 2018 23:07:20 +0200 Subject: [PATCH 22/98] drm/i915: Pass the new crtc_state to ->disable_plane() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit We're going to need access to the new crtc state in ->disable_plane() for SKL+ wm/ddb programming and pre-skl pipe gamma/csc control. Pass the crtc state down. We'll also try to make intel_crtc_disable_planes() do the right thing as much as it's possible. The fact that we don't have a separate crtc state for the disabled state when we're going to re-enable the crtc later means we might end up poking at a few extra planes in there. But that's harmless. I suppose one might argue that we wouldn't have to care about proper ddb/wm/csc/gamma if the pipe is going to permanently disable anyway, but the state checker probably cares so we should try our best to make sure everything is programmed correctly even in that case. v2: Fix the commit message a bit (Matt) Cc: Matt Roper Signed-off-by: Ville Syrjälä Link: https://patchwork.freedesktop.org/patch/msgid/20181114210729.16185-5-ville.syrjala@linux.intel.com Reviewed-by: Matt Roper Reviewed-by: Maarten Lankhorst --- drivers/gpu/drm/i915/intel_atomic_plane.c | 2 +- drivers/gpu/drm/i915/intel_display.c | 39 ++++++++++++++--------- drivers/gpu/drm/i915/intel_display.h | 8 +++++ drivers/gpu/drm/i915/intel_drv.h | 2 +- drivers/gpu/drm/i915/intel_sprite.c | 12 ++++--- 5 files changed, 42 insertions(+), 21 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_atomic_plane.c b/drivers/gpu/drm/i915/intel_atomic_plane.c index 026a11511941..ad3d135bdc15 100644 --- a/drivers/gpu/drm/i915/intel_atomic_plane.c +++ b/drivers/gpu/drm/i915/intel_atomic_plane.c @@ -212,7 +212,7 @@ void intel_update_planes_on_crtc(struct intel_atomic_state *old_state, } else { trace_intel_disable_plane(&plane->base, crtc); - plane->disable_plane(plane, crtc); + plane->disable_plane(plane, new_crtc_state); } } } diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 4469b4511c92..d088ffd9f542 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -2767,7 +2767,7 @@ static void intel_plane_disable_noatomic(struct intel_crtc *crtc, intel_pre_disable_primary_noatomic(&crtc->base); trace_intel_disable_plane(&plane->base, crtc); - plane->disable_plane(plane, crtc); + plane->disable_plane(plane, crtc_state); } static void @@ -3373,7 +3373,7 @@ static void i9xx_update_plane(struct intel_plane *plane, } static void i9xx_disable_plane(struct intel_plane *plane, - struct intel_crtc *crtc) + const struct intel_crtc_state *crtc_state) { struct drm_i915_private *dev_priv = to_i915(plane->base.dev); enum i9xx_plane_id i9xx_plane = plane->i9xx_plane; @@ -5406,23 +5406,32 @@ static void intel_pre_plane_update(struct intel_crtc_state *old_crtc_state, intel_update_watermarks(crtc); } -static void intel_crtc_disable_planes(struct intel_crtc *crtc, unsigned plane_mask) +static void intel_crtc_disable_planes(struct intel_atomic_state *state, + struct intel_crtc *crtc) { - struct drm_device *dev = crtc->base.dev; + struct drm_i915_private *dev_priv = to_i915(crtc->base.dev); + const struct intel_crtc_state *new_crtc_state = + intel_atomic_get_new_crtc_state(state, crtc); + unsigned int update_mask = new_crtc_state->update_planes; + const struct intel_plane_state *old_plane_state; struct intel_plane *plane; unsigned fb_bits = 0; + int i; intel_crtc_dpms_overlay_disable(crtc); - for_each_intel_plane_on_crtc(dev, crtc, plane) { - if (plane_mask & BIT(plane->id)) { - plane->disable_plane(plane, crtc); + for_each_old_intel_plane_in_state(state, plane, old_plane_state, i) { + if (crtc->pipe != plane->pipe || + !(update_mask & BIT(plane->id))) + continue; + plane->disable_plane(plane, new_crtc_state); + + if (old_plane_state->base.visible) fb_bits |= plane->frontbuffer_bit; - } } - intel_frontbuffer_flip(to_i915(dev), fb_bits); + intel_frontbuffer_flip(dev_priv, fb_bits); } static void intel_encoders_pre_pll_enable(struct drm_crtc *crtc, @@ -9897,9 +9906,9 @@ static void i845_update_cursor(struct intel_plane *plane, } static void i845_disable_cursor(struct intel_plane *plane, - struct intel_crtc *crtc) + const struct intel_crtc_state *crtc_state) { - i845_update_cursor(plane, NULL, NULL); + i845_update_cursor(plane, crtc_state, NULL); } static bool i845_cursor_get_hw_state(struct intel_plane *plane, @@ -10126,9 +10135,9 @@ static void i9xx_update_cursor(struct intel_plane *plane, } static void i9xx_disable_cursor(struct intel_plane *plane, - struct intel_crtc *crtc) + const struct intel_crtc_state *crtc_state) { - i9xx_update_cursor(plane, NULL, NULL); + i9xx_update_cursor(plane, crtc_state, NULL); } static bool i9xx_cursor_get_hw_state(struct intel_plane *plane, @@ -12892,7 +12901,7 @@ static void intel_atomic_commit_tail(struct drm_atomic_state *state) intel_pre_plane_update(old_intel_crtc_state, new_intel_crtc_state); if (old_crtc_state->active) { - intel_crtc_disable_planes(intel_crtc, old_intel_crtc_state->active_planes); + intel_crtc_disable_planes(intel_state, intel_crtc); /* * We need to disable pipe CRC before disabling the pipe, @@ -13742,7 +13751,7 @@ intel_legacy_cursor_update(struct drm_plane *plane, to_intel_plane_state(plane->state)); } else { trace_intel_disable_plane(plane, to_intel_crtc(crtc)); - intel_plane->disable_plane(intel_plane, to_intel_crtc(crtc)); + intel_plane->disable_plane(intel_plane, crtc_state); } intel_plane_unpin_fb(to_intel_plane_state(old_plane_state)); diff --git a/drivers/gpu/drm/i915/intel_display.h b/drivers/gpu/drm/i915/intel_display.h index 5f2955b944da..a9a181aa4587 100644 --- a/drivers/gpu/drm/i915/intel_display.h +++ b/drivers/gpu/drm/i915/intel_display.h @@ -398,6 +398,14 @@ struct intel_link_m_n { for_each_power_well_reverse(__dev_priv, __power_well) \ for_each_if((__power_well)->desc->domains & (__domain_mask)) +#define for_each_old_intel_plane_in_state(__state, plane, old_plane_state, __i) \ + for ((__i) = 0; \ + (__i) < (__state)->base.dev->mode_config.num_total_plane && \ + ((plane) = to_intel_plane((__state)->base.planes[__i].ptr), \ + (old_plane_state) = to_intel_plane_state((__state)->base.planes[__i].old_state), 1); \ + (__i)++) \ + for_each_if(plane) + #define for_each_new_intel_plane_in_state(__state, plane, new_plane_state, __i) \ for ((__i) = 0; \ (__i) < (__state)->base.dev->mode_config.num_total_plane && \ diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index 1f29061cefa1..7842d193ac44 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -1016,7 +1016,7 @@ struct intel_plane { const struct intel_crtc_state *crtc_state, const struct intel_plane_state *plane_state); void (*disable_plane)(struct intel_plane *plane, - struct intel_crtc *crtc); + const struct intel_crtc_state *crtc_state); bool (*get_hw_state)(struct intel_plane *plane, enum pipe *pipe); int (*check_plane)(struct intel_crtc_state *crtc_state, struct intel_plane_state *plane_state); diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c index 0548b996693f..e5da2fe3bdae 100644 --- a/drivers/gpu/drm/i915/intel_sprite.c +++ b/drivers/gpu/drm/i915/intel_sprite.c @@ -594,7 +594,8 @@ icl_update_slave(struct intel_plane *plane, } static void -skl_disable_plane(struct intel_plane *plane, struct intel_crtc *crtc) +skl_disable_plane(struct intel_plane *plane, + const struct intel_crtc_state *crtc_state) { struct drm_i915_private *dev_priv = to_i915(plane->base.dev); enum plane_id plane_id = plane->id; @@ -857,7 +858,8 @@ vlv_update_plane(struct intel_plane *plane, } static void -vlv_disable_plane(struct intel_plane *plane, struct intel_crtc *crtc) +vlv_disable_plane(struct intel_plane *plane, + const struct intel_crtc_state *crtc_state) { struct drm_i915_private *dev_priv = to_i915(plane->base.dev); enum pipe pipe = plane->pipe; @@ -1024,7 +1026,8 @@ ivb_update_plane(struct intel_plane *plane, } static void -ivb_disable_plane(struct intel_plane *plane, struct intel_crtc *crtc) +ivb_disable_plane(struct intel_plane *plane, + const struct intel_crtc_state *crtc_state) { struct drm_i915_private *dev_priv = to_i915(plane->base.dev); enum pipe pipe = plane->pipe; @@ -1190,7 +1193,8 @@ g4x_update_plane(struct intel_plane *plane, } static void -g4x_disable_plane(struct intel_plane *plane, struct intel_crtc *crtc) +g4x_disable_plane(struct intel_plane *plane, + const struct intel_crtc_state *crtc_state) { struct drm_i915_private *dev_priv = to_i915(plane->base.dev); enum pipe pipe = plane->pipe; From ce110ec311e9feae0905c24078146a2adbca877e Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Wed, 14 Nov 2018 23:07:21 +0200 Subject: [PATCH 23/98] drm/i915: Fix latency==0 handling for level 0 watermark on skl+ MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit If the level 0 latency is 0 we can't do anything. Return an error rather than success. While this can't happen due to WaWmMemoryReadLatency, it can happen if the user clears out the level 0 latency via debugfs. v2: Clarify how how we can end here with zero level 0 latency (Matt) Cc: Matt Roper Reviewed-by: Rodrigo Vivi Signed-off-by: Ville Syrjälä Link: https://patchwork.freedesktop.org/patch/msgid/20181114210729.16185-6-ville.syrjala@linux.intel.com Reviewed-by: Matt Roper Reviewed-by: Maarten Lankhorst --- drivers/gpu/drm/i915/intel_pm.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 897a791662c5..d94de52a8a76 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -4743,8 +4743,10 @@ static int skl_compute_plane_wm(const struct drm_i915_private *dev_priv, bool apply_memory_bw_wa = skl_needs_memory_bw_wa(state); uint32_t min_disp_buf_needed; - if (latency == 0 || - !intel_wm_plane_visible(cstate, intel_pstate)) { + if (latency == 0) + return level == 0 ? -EINVAL : 0; + + if (!intel_wm_plane_visible(cstate, intel_pstate)) { result->plane_en = false; return 0; } From 14a43062b90383f2cb0e634ff31f0bb7fa02cd1c Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Wed, 14 Nov 2018 23:07:22 +0200 Subject: [PATCH 24/98] drm/i915: Remove some useless zeroing on skl+ wm calculations MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit We memset(0) the entire watermark struct the start, so there's no need to clear things later on. v2: Rebase due to some stale w/a removal Reviewed-by: Rodrigo Vivi Signed-off-by: Ville Syrjälä Link: https://patchwork.freedesktop.org/patch/msgid/20181114210729.16185-7-ville.syrjala@linux.intel.com Reviewed-by: Matt Roper Reviewed-by: Maarten Lankhorst --- drivers/gpu/drm/i915/intel_pm.c | 16 ++++------------ 1 file changed, 4 insertions(+), 12 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index d94de52a8a76..54bf619b3185 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -4746,10 +4746,8 @@ static int skl_compute_plane_wm(const struct drm_i915_private *dev_priv, if (latency == 0) return level == 0 ? -EINVAL : 0; - if (!intel_wm_plane_visible(cstate, intel_pstate)) { - result->plane_en = false; + if (!intel_wm_plane_visible(cstate, intel_pstate)) return 0; - } /* Display WA #1141: kbl,cfl */ if ((IS_KABYLAKE(dev_priv) || IS_COFFEELAKE(dev_priv) || @@ -4846,8 +4844,6 @@ static int skl_compute_plane_wm(const struct drm_i915_private *dev_priv, if ((level > 0 && res_lines > 31) || res_blocks >= ddb_allocation || min_disp_buf_needed >= ddb_allocation) { - result->plane_en = false; - /* * If there are no valid level 0 watermarks, then we can't * support this display configuration. @@ -4949,15 +4945,15 @@ static void skl_compute_transition_wm(const struct intel_crtc_state *cstate, uint16_t wm0_sel_res_b, trans_offset_b, res_blocks; if (!cstate->base.active) - goto exit; + return; /* Transition WM are not recommended by HW team for GEN9 */ if (INTEL_GEN(dev_priv) <= 9) - goto exit; + return; /* Transition WM don't make any sense if ipc is disabled */ if (!dev_priv->ipc_enabled) - goto exit; + return; trans_min = 14; if (INTEL_GEN(dev_priv) >= 11) @@ -4996,11 +4992,7 @@ static void skl_compute_transition_wm(const struct intel_crtc_state *cstate, if (res_blocks < ddb_allocation) { trans_wm->plane_res_b = res_blocks; trans_wm->plane_en = true; - return; } - -exit: - trans_wm->plane_en = false; } static int __skl_build_plane_wm_single(struct skl_ddb_allocation *ddb, From 6a3c910b081d90fc7fddf771b249ff0059d5005c Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Wed, 14 Nov 2018 23:07:23 +0200 Subject: [PATCH 25/98] drm/i915: Pass the entire skl_plane_wm to skl_compute_transition_wm() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit We have to pass both level 0 watermark struct and the transition watermark struct to skl_compute_transition_wm(). Make life less confusing by just passing the entire plane watermark struct that contains both aforementioned structures. Cc: Matt Roper Signed-off-by: Ville Syrjälä Link: https://patchwork.freedesktop.org/patch/msgid/20181114210729.16185-8-ville.syrjala@linux.intel.com Reviewed-by: Matt Roper Reviewed-by: Maarten Lankhorst --- drivers/gpu/drm/i915/intel_pm.c | 16 +++++++--------- 1 file changed, 7 insertions(+), 9 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 54bf619b3185..db28fc547c65 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -4933,10 +4933,9 @@ skl_compute_linetime_wm(const struct intel_crtc_state *cstate) } static void skl_compute_transition_wm(const struct intel_crtc_state *cstate, - struct skl_wm_params *wp, - struct skl_wm_level *wm_l0, - uint16_t ddb_allocation, - struct skl_wm_level *trans_wm /* out */) + const struct skl_wm_params *wp, + struct skl_plane_wm *wm, + uint16_t ddb_allocation) { struct drm_device *dev = cstate->base.crtc->dev; const struct drm_i915_private *dev_priv = to_i915(dev); @@ -4971,7 +4970,7 @@ static void skl_compute_transition_wm(const struct intel_crtc_state *cstate, * Result Blocks is Result Blocks minus 1 and it should work for the * current platforms. */ - wm0_sel_res_b = wm_l0->plane_res_b - 1; + wm0_sel_res_b = wm->wm[0].plane_res_b - 1; if (wp->y_tiled) { trans_y_tile_min = (uint16_t) mul_round_up_u32_fixed16(2, @@ -4990,8 +4989,8 @@ static void skl_compute_transition_wm(const struct intel_crtc_state *cstate, res_blocks += 1; if (res_blocks < ddb_allocation) { - trans_wm->plane_res_b = res_blocks; - trans_wm->plane_en = true; + wm->trans_wm.plane_res_b = res_blocks; + wm->trans_wm.plane_en = true; } } @@ -5020,8 +5019,7 @@ static int __skl_build_plane_wm_single(struct skl_ddb_allocation *ddb, if (ret) return ret; - skl_compute_transition_wm(cstate, &wm_params, &wm->wm[0], - ddb_blocks, &wm->trans_wm); + skl_compute_transition_wm(cstate, &wm_params, wm, ddb_blocks); return 0; } From 8315847bf4df9290042f0d8bce3b1a2d303afe80 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Tue, 27 Nov 2018 18:57:26 +0200 Subject: [PATCH 26/98] drm/i915: Clean up skl+ vs. icl+ watermark computation MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Make a cleaner split between the skl+ and icl+ ways of computing watermarks. This way skl_build_pipe_wm() doesn't have to know any of the gritty details of icl+ master/slave planes. We can also simplify a bunch of the lower level code by pulling the plane visibility checks a bit higher up. v2: WARN_ON(!visible) for the icl+ master plane case (Matt) Cc: Matt Roper Reviewed-by: Maarten Lankhorst Reviewed-by: Matt Roper Signed-off-by: Ville Syrjälä Link: https://patchwork.freedesktop.org/patch/msgid/20181127165726.31122-1-ville.syrjala@linux.intel.com --- drivers/gpu/drm/i915/intel_pm.c | 197 +++++++++++++++++--------------- 1 file changed, 106 insertions(+), 91 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index db28fc547c65..99f825d1c2e7 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -4630,9 +4630,6 @@ skl_compute_plane_wm_params(const struct drm_i915_private *dev_priv, to_intel_atomic_state(cstate->base.state); bool apply_memory_bw_wa = skl_needs_memory_bw_wa(state); - if (!intel_wm_plane_visible(cstate, intel_pstate)) - return 0; - /* only NV12 format has two planes */ if (plane_id == 1 && fb->format->format != DRM_FORMAT_NV12) { DRM_DEBUG_KMS("Non NV12 format have single plane\n"); @@ -4746,9 +4743,6 @@ static int skl_compute_plane_wm(const struct drm_i915_private *dev_priv, if (latency == 0) return level == 0 ? -EINVAL : 0; - if (!intel_wm_plane_visible(cstate, intel_pstate)) - return 0; - /* Display WA #1141: kbl,cfl */ if ((IS_KABYLAKE(dev_priv) || IS_COFFEELAKE(dev_priv) || IS_CNL_REVID(dev_priv, CNL_REVID_A0, CNL_REVID_B0)) && @@ -4871,21 +4865,16 @@ static int skl_compute_plane_wm(const struct drm_i915_private *dev_priv, static int skl_compute_wm_levels(const struct drm_i915_private *dev_priv, - struct skl_ddb_allocation *ddb, const struct intel_crtc_state *cstate, const struct intel_plane_state *intel_pstate, uint16_t ddb_blocks, const struct skl_wm_params *wm_params, - struct skl_plane_wm *wm, struct skl_wm_level *levels) { int level, max_level = ilk_wm_max_level(dev_priv); struct skl_wm_level *result_prev = &levels[0]; int ret; - if (WARN_ON(!intel_pstate->base.fb)) - return -EINVAL; - for (level = 0; level <= max_level; level++) { struct skl_wm_level *result = &levels[level]; @@ -4903,9 +4892,6 @@ skl_compute_wm_levels(const struct drm_i915_private *dev_priv, result_prev = result; } - if (intel_pstate->base.fb->format->format == DRM_FORMAT_NV12) - wm->is_planar = true; - return 0; } @@ -4943,9 +4929,6 @@ static void skl_compute_transition_wm(const struct intel_crtc_state *cstate, const uint16_t trans_amount = 10; /* This is configurable amount */ uint16_t wm0_sel_res_b, trans_offset_b, res_blocks; - if (!cstate->base.active) - return; - /* Transition WM are not recommended by HW team for GEN9 */ if (INTEL_GEN(dev_priv) <= 9) return; @@ -4994,97 +4977,135 @@ static void skl_compute_transition_wm(const struct intel_crtc_state *cstate, } } -static int __skl_build_plane_wm_single(struct skl_ddb_allocation *ddb, - struct skl_pipe_wm *pipe_wm, - enum plane_id plane_id, - const struct intel_crtc_state *cstate, - const struct intel_plane_state *pstate, - int color_plane) -{ - struct drm_i915_private *dev_priv = to_i915(pstate->base.plane->dev); - struct skl_plane_wm *wm = &pipe_wm->planes[plane_id]; - enum pipe pipe = to_intel_plane(pstate->base.plane)->pipe; - struct skl_wm_params wm_params; - uint16_t ddb_blocks = skl_ddb_entry_size(&ddb->plane[pipe][plane_id]); - int ret; - - ret = skl_compute_plane_wm_params(dev_priv, cstate, pstate, - &wm_params, color_plane); - if (ret) - return ret; - - ret = skl_compute_wm_levels(dev_priv, ddb, cstate, pstate, - ddb_blocks, &wm_params, wm, wm->wm); - - if (ret) - return ret; - - skl_compute_transition_wm(cstate, &wm_params, wm, ddb_blocks); - - return 0; -} - static int skl_build_plane_wm_single(struct skl_ddb_allocation *ddb, - struct skl_pipe_wm *pipe_wm, - const struct intel_crtc_state *cstate, - const struct intel_plane_state *pstate) + struct intel_crtc_state *crtc_state, + const struct intel_plane_state *plane_state, + enum plane_id plane_id, int color_plane) { - enum plane_id plane_id = to_intel_plane(pstate->base.plane)->id; - - return __skl_build_plane_wm_single(ddb, pipe_wm, plane_id, cstate, pstate, 0); -} - -static int skl_build_plane_wm_planar(struct skl_ddb_allocation *ddb, - struct skl_pipe_wm *pipe_wm, - const struct intel_crtc_state *cstate, - const struct intel_plane_state *pstate) -{ - struct intel_plane *plane = to_intel_plane(pstate->base.plane); + struct intel_plane *plane = to_intel_plane(plane_state->base.plane); struct drm_i915_private *dev_priv = to_i915(plane->base.dev); - enum plane_id plane_id = plane->id; - struct skl_plane_wm *wm = &pipe_wm->planes[plane_id]; + struct skl_plane_wm *wm = &crtc_state->wm.skl.optimal.planes[plane_id]; struct skl_wm_params wm_params; enum pipe pipe = plane->pipe; uint16_t ddb_blocks = skl_ddb_entry_size(&ddb->plane[pipe][plane_id]); int ret; - ret = __skl_build_plane_wm_single(ddb, pipe_wm, plane_id, cstate, pstate, 0); + ret = skl_compute_plane_wm_params(dev_priv, crtc_state, plane_state, + &wm_params, color_plane); if (ret) return ret; - /* uv plane watermarks must also be validated for NV12/Planar */ - ddb_blocks = skl_ddb_entry_size(&ddb->uv_plane[pipe][plane_id]); - - ret = skl_compute_plane_wm_params(dev_priv, cstate, pstate, &wm_params, 1); + ret = skl_compute_wm_levels(dev_priv, crtc_state, plane_state, + ddb_blocks, &wm_params, wm->wm); if (ret) return ret; - return skl_compute_wm_levels(dev_priv, ddb, cstate, pstate, - ddb_blocks, &wm_params, wm, wm->uv_wm); + skl_compute_transition_wm(crtc_state, &wm_params, wm, ddb_blocks); + + return 0; } -static int icl_build_plane_wm_planar(struct skl_ddb_allocation *ddb, - struct skl_pipe_wm *pipe_wm, - const struct intel_crtc_state *cstate, - const struct intel_plane_state *pstate) +static int skl_build_plane_wm_uv(struct skl_ddb_allocation *ddb, + struct intel_crtc_state *crtc_state, + const struct intel_plane_state *plane_state, + enum plane_id plane_id) { + struct intel_plane *plane = to_intel_plane(plane_state->base.plane); + struct drm_i915_private *dev_priv = to_i915(plane->base.dev); + struct skl_plane_wm *wm = &crtc_state->wm.skl.optimal.planes[plane_id]; + struct skl_wm_params wm_params; + enum pipe pipe = plane->pipe; + uint16_t ddb_blocks = skl_ddb_entry_size(&ddb->uv_plane[pipe][plane_id]); int ret; - enum plane_id y_plane_id = pstate->linked_plane->id; - enum plane_id uv_plane_id = to_intel_plane(pstate->base.plane)->id; - ret = __skl_build_plane_wm_single(ddb, pipe_wm, y_plane_id, - cstate, pstate, 0); + wm->is_planar = true; + + /* uv plane watermarks must also be validated for NV12/Planar */ + ret = skl_compute_plane_wm_params(dev_priv, crtc_state, plane_state, + &wm_params, 1); if (ret) return ret; - return __skl_build_plane_wm_single(ddb, pipe_wm, uv_plane_id, - cstate, pstate, 1); + ret = skl_compute_wm_levels(dev_priv, crtc_state, plane_state, + ddb_blocks, &wm_params, wm->uv_wm); + if (ret) + return ret; + + return 0; +} + +static int skl_build_plane_wm(struct skl_ddb_allocation *ddb, + struct skl_pipe_wm *pipe_wm, + struct intel_crtc_state *crtc_state, + const struct intel_plane_state *plane_state) +{ + struct intel_plane *plane = to_intel_plane(plane_state->base.plane); + const struct drm_framebuffer *fb = plane_state->base.fb; + enum plane_id plane_id = plane->id; + int ret; + + if (!intel_wm_plane_visible(crtc_state, plane_state)) + return 0; + + ret = skl_build_plane_wm_single(ddb, crtc_state, plane_state, + plane_id, 0); + if (ret) + return ret; + + if (fb->format->is_yuv && fb->format->num_planes > 1) { + ret = skl_build_plane_wm_uv(ddb, crtc_state, plane_state, + plane_id); + if (ret) + return ret; + } + + return 0; +} + +static int icl_build_plane_wm(struct skl_ddb_allocation *ddb, + struct skl_pipe_wm *pipe_wm, + struct intel_crtc_state *crtc_state, + const struct intel_plane_state *plane_state) +{ + enum plane_id plane_id = to_intel_plane(plane_state->base.plane)->id; + int ret; + + /* Watermarks calculated in master */ + if (plane_state->slave) + return 0; + + if (plane_state->linked_plane) { + const struct drm_framebuffer *fb = plane_state->base.fb; + enum plane_id y_plane_id = plane_state->linked_plane->id; + + WARN_ON(!intel_wm_plane_visible(crtc_state, plane_state)); + WARN_ON(!fb->format->is_yuv || + fb->format->num_planes == 1); + + ret = skl_build_plane_wm_single(ddb, crtc_state, plane_state, + y_plane_id, 0); + if (ret) + return ret; + + ret = skl_build_plane_wm_single(ddb, crtc_state, plane_state, + plane_id, 1); + if (ret) + return ret; + } else if (intel_wm_plane_visible(crtc_state, plane_state)) { + ret = skl_build_plane_wm_single(ddb, crtc_state, plane_state, + plane_id, 0); + if (ret) + return ret; + } + + return 0; } static int skl_build_pipe_wm(struct intel_crtc_state *cstate, struct skl_ddb_allocation *ddb, struct skl_pipe_wm *pipe_wm) { + struct drm_i915_private *dev_priv = to_i915(cstate->base.crtc->dev); struct drm_crtc_state *crtc_state = &cstate->base; struct drm_plane *plane; const struct drm_plane_state *pstate; @@ -5100,18 +5121,12 @@ static int skl_build_pipe_wm(struct intel_crtc_state *cstate, const struct intel_plane_state *intel_pstate = to_intel_plane_state(pstate); - /* Watermarks calculated in master */ - if (intel_pstate->slave) - continue; - - if (intel_pstate->linked_plane) - ret = icl_build_plane_wm_planar(ddb, pipe_wm, cstate, intel_pstate); - else if (intel_pstate->base.fb && - intel_pstate->base.fb->format->format == DRM_FORMAT_NV12) - ret = skl_build_plane_wm_planar(ddb, pipe_wm, cstate, intel_pstate); + if (INTEL_GEN(dev_priv) >= 11) + ret = icl_build_plane_wm(ddb, pipe_wm, + cstate, intel_pstate); else - ret = skl_build_plane_wm_single(ddb, pipe_wm, cstate, intel_pstate); - + ret = skl_build_plane_wm(ddb, pipe_wm, + cstate, intel_pstate); if (ret) return ret; } From 51de9c6d25594f70c0a03466546ca3deb9705d0e Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Wed, 14 Nov 2018 23:07:25 +0200 Subject: [PATCH 27/98] drm/i915: Don't pass dev_priv around so much MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Simplify the calling convention of the skl+ watermark functions by not passing around dev_priv needlessly. The callees have what they need to dig it out anyway. Signed-off-by: Ville Syrjälä Link: https://patchwork.freedesktop.org/patch/msgid/20181114210729.16185-10-ville.syrjala@linux.intel.com Reviewed-by: Matt Roper Reviewed-by: Maarten Lankhorst --- drivers/gpu/drm/i915/intel_pm.c | 27 +++++++++++++-------------- 1 file changed, 13 insertions(+), 14 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 99f825d1c2e7..a5891f7a9f9d 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -4617,12 +4617,12 @@ skl_adjusted_plane_pixel_rate(const struct intel_crtc_state *cstate, } static int -skl_compute_plane_wm_params(const struct drm_i915_private *dev_priv, - const struct intel_crtc_state *cstate, +skl_compute_plane_wm_params(const struct intel_crtc_state *cstate, const struct intel_plane_state *intel_pstate, struct skl_wm_params *wp, int plane_id) { struct intel_plane *plane = to_intel_plane(intel_pstate->base.plane); + struct drm_i915_private *dev_priv = to_i915(plane->base.dev); const struct drm_plane_state *pstate = &intel_pstate->base; const struct drm_framebuffer *fb = pstate->fb; uint32_t interm_pbpl; @@ -4721,8 +4721,7 @@ skl_compute_plane_wm_params(const struct drm_i915_private *dev_priv, return 0; } -static int skl_compute_plane_wm(const struct drm_i915_private *dev_priv, - const struct intel_crtc_state *cstate, +static int skl_compute_plane_wm(const struct intel_crtc_state *cstate, const struct intel_plane_state *intel_pstate, uint16_t ddb_allocation, int level, @@ -4730,6 +4729,8 @@ static int skl_compute_plane_wm(const struct drm_i915_private *dev_priv, const struct skl_wm_level *result_prev, struct skl_wm_level *result /* out */) { + struct drm_i915_private *dev_priv = + to_i915(intel_pstate->base.plane->dev); const struct drm_plane_state *pstate = &intel_pstate->base; uint32_t latency = dev_priv->wm.skl_latency[level]; uint_fixed_16_16_t method1, method2; @@ -4864,13 +4865,14 @@ static int skl_compute_plane_wm(const struct drm_i915_private *dev_priv, } static int -skl_compute_wm_levels(const struct drm_i915_private *dev_priv, - const struct intel_crtc_state *cstate, +skl_compute_wm_levels(const struct intel_crtc_state *cstate, const struct intel_plane_state *intel_pstate, uint16_t ddb_blocks, const struct skl_wm_params *wm_params, struct skl_wm_level *levels) { + struct drm_i915_private *dev_priv = + to_i915(intel_pstate->base.plane->dev); int level, max_level = ilk_wm_max_level(dev_priv); struct skl_wm_level *result_prev = &levels[0]; int ret; @@ -4878,8 +4880,7 @@ skl_compute_wm_levels(const struct drm_i915_private *dev_priv, for (level = 0; level <= max_level; level++) { struct skl_wm_level *result = &levels[level]; - ret = skl_compute_plane_wm(dev_priv, - cstate, + ret = skl_compute_plane_wm(cstate, intel_pstate, ddb_blocks, level, @@ -4983,19 +4984,18 @@ static int skl_build_plane_wm_single(struct skl_ddb_allocation *ddb, enum plane_id plane_id, int color_plane) { struct intel_plane *plane = to_intel_plane(plane_state->base.plane); - struct drm_i915_private *dev_priv = to_i915(plane->base.dev); struct skl_plane_wm *wm = &crtc_state->wm.skl.optimal.planes[plane_id]; struct skl_wm_params wm_params; enum pipe pipe = plane->pipe; uint16_t ddb_blocks = skl_ddb_entry_size(&ddb->plane[pipe][plane_id]); int ret; - ret = skl_compute_plane_wm_params(dev_priv, crtc_state, plane_state, + ret = skl_compute_plane_wm_params(crtc_state, plane_state, &wm_params, color_plane); if (ret) return ret; - ret = skl_compute_wm_levels(dev_priv, crtc_state, plane_state, + ret = skl_compute_wm_levels(crtc_state, plane_state, ddb_blocks, &wm_params, wm->wm); if (ret) return ret; @@ -5011,7 +5011,6 @@ static int skl_build_plane_wm_uv(struct skl_ddb_allocation *ddb, enum plane_id plane_id) { struct intel_plane *plane = to_intel_plane(plane_state->base.plane); - struct drm_i915_private *dev_priv = to_i915(plane->base.dev); struct skl_plane_wm *wm = &crtc_state->wm.skl.optimal.planes[plane_id]; struct skl_wm_params wm_params; enum pipe pipe = plane->pipe; @@ -5021,12 +5020,12 @@ static int skl_build_plane_wm_uv(struct skl_ddb_allocation *ddb, wm->is_planar = true; /* uv plane watermarks must also be validated for NV12/Planar */ - ret = skl_compute_plane_wm_params(dev_priv, crtc_state, plane_state, + ret = skl_compute_plane_wm_params(crtc_state, plane_state, &wm_params, 1); if (ret) return ret; - ret = skl_compute_wm_levels(dev_priv, crtc_state, plane_state, + ret = skl_compute_wm_levels(crtc_state, plane_state, ddb_blocks, &wm_params, wm->uv_wm); if (ret) return ret; From ff43bc379e16c9195323cb88ac0c9f4d0613d07a Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Tue, 27 Nov 2018 18:59:00 +0200 Subject: [PATCH 28/98] drm/i915: Move ddb/wm programming into plane update/disable hooks on skl+ MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit On SKL+ the plane WM/BUF_CFG registers are a proper part of each plane's register set. That means accessing them will cancel any pending plane update, and we would need a PLANE_SURF register write to arm the wm/ddb change as well. To avoid all the problems with that let's just move the wm/ddb programming into the plane update/disable hooks. Now all plane registers get written in one (hopefully atomic) operation. To make that feasible we'll move the plane ddb tracking into the crtc state. Watermarks were already tracked there. v2: Rebase due to input CSC v3: Split out a bunch of junk (Matt) v4: Add skl_wm_add_affected_planes() to deal with cursor special case and non-zero wm register reset value v5: Drop the unrelated for_each_intel_plane_mask() fix (Matt) Remove the redundant ddb memset() (Matt) Cc: Matt Roper Cc: Maarten Lankhorst Reviewed-by: Maarten Lankhorst #v3 Signed-off-by: Ville Syrjälä Link: https://patchwork.freedesktop.org/patch/msgid/20181127165900.31298-1-ville.syrjala@linux.intel.com --- drivers/gpu/drm/i915/i915_debugfs.c | 21 +- drivers/gpu/drm/i915/i915_drv.h | 3 - drivers/gpu/drm/i915/intel_display.c | 16 +- drivers/gpu/drm/i915/intel_display.h | 9 + drivers/gpu/drm/i915/intel_drv.h | 9 + drivers/gpu/drm/i915/intel_pm.c | 407 ++++++++++++++------------- drivers/gpu/drm/i915/intel_sprite.c | 4 + 7 files changed, 261 insertions(+), 208 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c index 596810e0cfe8..129b9a6f8309 100644 --- a/drivers/gpu/drm/i915/i915_debugfs.c +++ b/drivers/gpu/drm/i915/i915_debugfs.c @@ -3441,31 +3441,32 @@ static int i915_ddb_info(struct seq_file *m, void *unused) { struct drm_i915_private *dev_priv = node_to_i915(m->private); struct drm_device *dev = &dev_priv->drm; - struct skl_ddb_allocation *ddb; struct skl_ddb_entry *entry; - enum pipe pipe; - int plane; + struct intel_crtc *crtc; if (INTEL_GEN(dev_priv) < 9) return -ENODEV; drm_modeset_lock_all(dev); - ddb = &dev_priv->wm.skl_hw.ddb; - seq_printf(m, "%-15s%8s%8s%8s\n", "", "Start", "End", "Size"); - for_each_pipe(dev_priv, pipe) { + for_each_intel_crtc(&dev_priv->drm, crtc) { + struct intel_crtc_state *crtc_state = + to_intel_crtc_state(crtc->base.state); + enum pipe pipe = crtc->pipe; + enum plane_id plane_id; + seq_printf(m, "Pipe %c\n", pipe_name(pipe)); - for_each_universal_plane(dev_priv, pipe, plane) { - entry = &ddb->plane[pipe][plane]; - seq_printf(m, " Plane%-8d%8u%8u%8u\n", plane + 1, + for_each_plane_id_on_crtc(crtc, plane_id) { + entry = &crtc_state->wm.skl.plane_ddb_y[plane_id]; + seq_printf(m, " Plane%-8d%8u%8u%8u\n", plane_id + 1, entry->start, entry->end, skl_ddb_entry_size(entry)); } - entry = &ddb->plane[pipe][PLANE_CURSOR]; + entry = &crtc_state->wm.skl.plane_ddb_y[PLANE_CURSOR]; seq_printf(m, " %-13s%8u%8u%8u\n", "Cursor", entry->start, entry->end, skl_ddb_entry_size(entry)); } diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index f763b30f98d9..645c2bbdcdfa 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -1095,9 +1095,6 @@ static inline bool skl_ddb_entry_equal(const struct skl_ddb_entry *e1, } struct skl_ddb_allocation { - /* packed/y */ - struct skl_ddb_entry plane[I915_MAX_PIPES][I915_MAX_PLANES]; - struct skl_ddb_entry uv_plane[I915_MAX_PIPES][I915_MAX_PLANES]; u8 enabled_slices; /* GEN11 has configurable 2 slices */ }; diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index d088ffd9f542..6c5d71b44740 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -10114,6 +10114,10 @@ static void i9xx_update_cursor(struct intel_plane *plane, * except when the plane is getting enabled at which time * the CURCNTR write arms the update. */ + + if (INTEL_GEN(dev_priv) >= 9) + skl_write_cursor_wm(plane, crtc_state); + if (plane->cursor.base != base || plane->cursor.size != fbc_ctl || plane->cursor.cntl != cntl) { @@ -11903,6 +11907,8 @@ static void verify_wm_state(struct drm_crtc *crtc, struct skl_pipe_wm hw_wm, *sw_wm; struct skl_plane_wm *hw_plane_wm, *sw_plane_wm; struct skl_ddb_entry *hw_ddb_entry, *sw_ddb_entry; + struct skl_ddb_entry hw_ddb_y[I915_MAX_PLANES]; + struct skl_ddb_entry hw_ddb_uv[I915_MAX_PLANES]; struct intel_crtc *intel_crtc = to_intel_crtc(crtc); const enum pipe pipe = intel_crtc->pipe; int plane, level, max_level = ilk_wm_max_level(dev_priv); @@ -11913,6 +11919,8 @@ static void verify_wm_state(struct drm_crtc *crtc, skl_pipe_wm_get_hw_state(crtc, &hw_wm); sw_wm = &to_intel_crtc_state(new_state)->wm.skl.optimal; + skl_pipe_ddb_get_hw_state(intel_crtc, hw_ddb_y, hw_ddb_uv); + skl_ddb_get_hw_state(dev_priv, &hw_ddb); sw_ddb = &dev_priv->wm.skl_hw.ddb; @@ -11955,8 +11963,8 @@ static void verify_wm_state(struct drm_crtc *crtc, } /* DDB */ - hw_ddb_entry = &hw_ddb.plane[pipe][plane]; - sw_ddb_entry = &sw_ddb->plane[pipe][plane]; + hw_ddb_entry = &hw_ddb_y[plane]; + sw_ddb_entry = &to_intel_crtc_state(new_state)->wm.skl.plane_ddb_y[plane]; if (!skl_ddb_entry_equal(hw_ddb_entry, sw_ddb_entry)) { DRM_ERROR("mismatch in DDB state pipe %c plane %d (expected (%u,%u), found (%u,%u))\n", @@ -12005,8 +12013,8 @@ static void verify_wm_state(struct drm_crtc *crtc, } /* DDB */ - hw_ddb_entry = &hw_ddb.plane[pipe][PLANE_CURSOR]; - sw_ddb_entry = &sw_ddb->plane[pipe][PLANE_CURSOR]; + hw_ddb_entry = &hw_ddb_y[PLANE_CURSOR]; + sw_ddb_entry = &to_intel_crtc_state(new_state)->wm.skl.plane_ddb_y[PLANE_CURSOR]; if (!skl_ddb_entry_equal(hw_ddb_entry, sw_ddb_entry)) { DRM_ERROR("mismatch in DDB state pipe %c cursor (expected (%u,%u), found (%u,%u))\n", diff --git a/drivers/gpu/drm/i915/intel_display.h b/drivers/gpu/drm/i915/intel_display.h index a9a181aa4587..a7ceb8f904f7 100644 --- a/drivers/gpu/drm/i915/intel_display.h +++ b/drivers/gpu/drm/i915/intel_display.h @@ -431,6 +431,15 @@ struct intel_link_m_n { (__i)++) \ for_each_if(plane) +#define for_each_oldnew_intel_crtc_in_state(__state, crtc, old_crtc_state, new_crtc_state, __i) \ + for ((__i) = 0; \ + (__i) < (__state)->base.dev->mode_config.num_crtc && \ + ((crtc) = to_intel_crtc((__state)->base.crtcs[__i].ptr), \ + (old_crtc_state) = to_intel_crtc_state((__state)->base.crtcs[__i].old_state), \ + (new_crtc_state) = to_intel_crtc_state((__state)->base.crtcs[__i].new_state), 1); \ + (__i)++) \ + for_each_if(crtc) + void intel_link_compute_m_n(int bpp, int nlanes, int pixel_clock, int link_clock, struct intel_link_m_n *m_n, diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index 7842d193ac44..1a3a396862d1 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -706,6 +706,8 @@ struct intel_crtc_wm_state { /* gen9+ only needs 1-step wm programming */ struct skl_pipe_wm optimal; struct skl_ddb_entry ddb; + struct skl_ddb_entry plane_ddb_y[I915_MAX_PLANES]; + struct skl_ddb_entry plane_ddb_uv[I915_MAX_PLANES]; } skl; struct { @@ -2185,6 +2187,9 @@ void g4x_wm_get_hw_state(struct drm_device *dev); void vlv_wm_get_hw_state(struct drm_device *dev); void ilk_wm_get_hw_state(struct drm_device *dev); void skl_wm_get_hw_state(struct drm_device *dev); +void skl_pipe_ddb_get_hw_state(struct intel_crtc *crtc, + struct skl_ddb_entry *ddb_y, + struct skl_ddb_entry *ddb_uv); void skl_ddb_get_hw_state(struct drm_i915_private *dev_priv, struct skl_ddb_allocation *ddb /* out */); void skl_pipe_wm_get_hw_state(struct drm_crtc *crtc, @@ -2199,6 +2204,10 @@ bool skl_wm_level_equals(const struct skl_wm_level *l1, bool skl_ddb_allocation_overlaps(const struct skl_ddb_entry *ddb, const struct skl_ddb_entry entries[], int num_entries, int ignore_idx); +void skl_write_plane_wm(struct intel_plane *plane, + const struct intel_crtc_state *crtc_state); +void skl_write_cursor_wm(struct intel_plane *plane, + const struct intel_crtc_state *crtc_state); bool ilk_disable_lp_wm(struct drm_device *dev); int skl_check_pipe_max_pixel_rate(struct intel_crtc *intel_crtc, struct intel_crtc_state *cstate); diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index a5891f7a9f9d..b1f30d56b747 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -3951,68 +3951,68 @@ static void skl_ddb_get_hw_plane_state(struct drm_i915_private *dev_priv, const enum pipe pipe, const enum plane_id plane_id, - struct skl_ddb_allocation *ddb /* out */) + struct skl_ddb_entry *ddb_y, + struct skl_ddb_entry *ddb_uv) { - u32 val, val2 = 0; - int fourcc, pixel_format; + u32 val, val2; + u32 fourcc = 0; /* Cursor doesn't support NV12/planar, so no extra calculation needed */ if (plane_id == PLANE_CURSOR) { val = I915_READ(CUR_BUF_CFG(pipe)); - skl_ddb_entry_init_from_hw(dev_priv, - &ddb->plane[pipe][plane_id], val); + skl_ddb_entry_init_from_hw(dev_priv, ddb_y, val); return; } val = I915_READ(PLANE_CTL(pipe, plane_id)); /* No DDB allocated for disabled planes */ - if (!(val & PLANE_CTL_ENABLE)) - return; + if (val & PLANE_CTL_ENABLE) + fourcc = skl_format_to_fourcc(val & PLANE_CTL_FORMAT_MASK, + val & PLANE_CTL_ORDER_RGBX, + val & PLANE_CTL_ALPHA_MASK); - pixel_format = val & PLANE_CTL_FORMAT_MASK; - fourcc = skl_format_to_fourcc(pixel_format, - val & PLANE_CTL_ORDER_RGBX, - val & PLANE_CTL_ALPHA_MASK); - - val = I915_READ(PLANE_BUF_CFG(pipe, plane_id)); - if (fourcc == DRM_FORMAT_NV12 && INTEL_GEN(dev_priv) < 11) { + if (INTEL_GEN(dev_priv) >= 11) { + val = I915_READ(PLANE_BUF_CFG(pipe, plane_id)); + skl_ddb_entry_init_from_hw(dev_priv, ddb_y, val); + } else { + val = I915_READ(PLANE_BUF_CFG(pipe, plane_id)); val2 = I915_READ(PLANE_NV12_BUF_CFG(pipe, plane_id)); - skl_ddb_entry_init_from_hw(dev_priv, - &ddb->plane[pipe][plane_id], val2); - skl_ddb_entry_init_from_hw(dev_priv, - &ddb->uv_plane[pipe][plane_id], val); - } else { - skl_ddb_entry_init_from_hw(dev_priv, - &ddb->plane[pipe][plane_id], val); + if (fourcc == DRM_FORMAT_NV12) + swap(val, val2); + + skl_ddb_entry_init_from_hw(dev_priv, ddb_y, val); + skl_ddb_entry_init_from_hw(dev_priv, ddb_uv, val2); } } +void skl_pipe_ddb_get_hw_state(struct intel_crtc *crtc, + struct skl_ddb_entry *ddb_y, + struct skl_ddb_entry *ddb_uv) +{ + struct drm_i915_private *dev_priv = to_i915(crtc->base.dev); + enum intel_display_power_domain power_domain; + enum pipe pipe = crtc->pipe; + enum plane_id plane_id; + + power_domain = POWER_DOMAIN_PIPE(pipe); + if (!intel_display_power_get_if_enabled(dev_priv, power_domain)) + return; + + for_each_plane_id_on_crtc(crtc, plane_id) + skl_ddb_get_hw_plane_state(dev_priv, pipe, + plane_id, + &ddb_y[plane_id], + &ddb_uv[plane_id]); + + intel_display_power_put(dev_priv, power_domain); +} + void skl_ddb_get_hw_state(struct drm_i915_private *dev_priv, struct skl_ddb_allocation *ddb /* out */) { - struct intel_crtc *crtc; - - memset(ddb, 0, sizeof(*ddb)); - ddb->enabled_slices = intel_enabled_dbuf_slices_num(dev_priv); - - for_each_intel_crtc(&dev_priv->drm, crtc) { - enum intel_display_power_domain power_domain; - enum plane_id plane_id; - enum pipe pipe = crtc->pipe; - - power_domain = POWER_DOMAIN_PIPE(pipe); - if (!intel_display_power_get_if_enabled(dev_priv, power_domain)) - continue; - - for_each_plane_id_on_crtc(crtc, plane_id) - skl_ddb_get_hw_plane_state(dev_priv, pipe, - plane_id, ddb); - - intel_display_power_put(dev_priv, power_domain); - } } /* @@ -4410,7 +4410,6 @@ skl_allocate_pipe_ddb(struct intel_crtc_state *cstate, struct drm_crtc *crtc = cstate->base.crtc; struct drm_i915_private *dev_priv = to_i915(crtc->dev); struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - enum pipe pipe = intel_crtc->pipe; struct skl_ddb_entry *alloc = &cstate->wm.skl.ddb; uint16_t alloc_size, start; uint16_t minimum[I915_MAX_PLANES] = {}; @@ -4423,8 +4422,8 @@ skl_allocate_pipe_ddb(struct intel_crtc_state *cstate, uint16_t total_min_blocks = 0; /* Clear the partitioning for disabled planes. */ - memset(ddb->plane[pipe], 0, sizeof(ddb->plane[pipe])); - memset(ddb->uv_plane[pipe], 0, sizeof(ddb->uv_plane[pipe])); + memset(cstate->wm.skl.plane_ddb_y, 0, sizeof(cstate->wm.skl.plane_ddb_y)); + memset(cstate->wm.skl.plane_ddb_uv, 0, sizeof(cstate->wm.skl.plane_ddb_uv)); if (WARN_ON(!state)) return 0; @@ -4471,8 +4470,8 @@ skl_allocate_pipe_ddb(struct intel_crtc_state *cstate, } alloc_size -= total_min_blocks; - ddb->plane[pipe][PLANE_CURSOR].start = alloc->end - minimum[PLANE_CURSOR]; - ddb->plane[pipe][PLANE_CURSOR].end = alloc->end; + cstate->wm.skl.plane_ddb_y[PLANE_CURSOR].start = alloc->end - minimum[PLANE_CURSOR]; + cstate->wm.skl.plane_ddb_y[PLANE_CURSOR].end = alloc->end; /* * 2. Distribute the remaining space in proportion to the amount of @@ -4503,8 +4502,8 @@ skl_allocate_pipe_ddb(struct intel_crtc_state *cstate, /* Leave disabled planes at (0,0) */ if (data_rate) { - ddb->plane[pipe][plane_id].start = start; - ddb->plane[pipe][plane_id].end = start + plane_blocks; + cstate->wm.skl.plane_ddb_y[plane_id].start = start; + cstate->wm.skl.plane_ddb_y[plane_id].end = start + plane_blocks; } start += plane_blocks; @@ -4519,8 +4518,8 @@ skl_allocate_pipe_ddb(struct intel_crtc_state *cstate, WARN_ON(INTEL_GEN(dev_priv) >= 11 && uv_plane_blocks); if (uv_data_rate) { - ddb->uv_plane[pipe][plane_id].start = start; - ddb->uv_plane[pipe][plane_id].end = + cstate->wm.skl.plane_ddb_uv[plane_id].start = start; + cstate->wm.skl.plane_ddb_uv[plane_id].end = start + uv_plane_blocks; } @@ -4978,16 +4977,13 @@ static void skl_compute_transition_wm(const struct intel_crtc_state *cstate, } } -static int skl_build_plane_wm_single(struct skl_ddb_allocation *ddb, - struct intel_crtc_state *crtc_state, +static int skl_build_plane_wm_single(struct intel_crtc_state *crtc_state, const struct intel_plane_state *plane_state, enum plane_id plane_id, int color_plane) { - struct intel_plane *plane = to_intel_plane(plane_state->base.plane); struct skl_plane_wm *wm = &crtc_state->wm.skl.optimal.planes[plane_id]; + u16 ddb_blocks = skl_ddb_entry_size(&crtc_state->wm.skl.plane_ddb_y[plane_id]); struct skl_wm_params wm_params; - enum pipe pipe = plane->pipe; - uint16_t ddb_blocks = skl_ddb_entry_size(&ddb->plane[pipe][plane_id]); int ret; ret = skl_compute_plane_wm_params(crtc_state, plane_state, @@ -5005,16 +5001,13 @@ static int skl_build_plane_wm_single(struct skl_ddb_allocation *ddb, return 0; } -static int skl_build_plane_wm_uv(struct skl_ddb_allocation *ddb, - struct intel_crtc_state *crtc_state, +static int skl_build_plane_wm_uv(struct intel_crtc_state *crtc_state, const struct intel_plane_state *plane_state, enum plane_id plane_id) { - struct intel_plane *plane = to_intel_plane(plane_state->base.plane); struct skl_plane_wm *wm = &crtc_state->wm.skl.optimal.planes[plane_id]; + u16 ddb_blocks = skl_ddb_entry_size(&crtc_state->wm.skl.plane_ddb_uv[plane_id]); struct skl_wm_params wm_params; - enum pipe pipe = plane->pipe; - uint16_t ddb_blocks = skl_ddb_entry_size(&ddb->uv_plane[pipe][plane_id]); int ret; wm->is_planar = true; @@ -5033,8 +5026,7 @@ static int skl_build_plane_wm_uv(struct skl_ddb_allocation *ddb, return 0; } -static int skl_build_plane_wm(struct skl_ddb_allocation *ddb, - struct skl_pipe_wm *pipe_wm, +static int skl_build_plane_wm(struct skl_pipe_wm *pipe_wm, struct intel_crtc_state *crtc_state, const struct intel_plane_state *plane_state) { @@ -5046,13 +5038,13 @@ static int skl_build_plane_wm(struct skl_ddb_allocation *ddb, if (!intel_wm_plane_visible(crtc_state, plane_state)) return 0; - ret = skl_build_plane_wm_single(ddb, crtc_state, plane_state, + ret = skl_build_plane_wm_single(crtc_state, plane_state, plane_id, 0); if (ret) return ret; if (fb->format->is_yuv && fb->format->num_planes > 1) { - ret = skl_build_plane_wm_uv(ddb, crtc_state, plane_state, + ret = skl_build_plane_wm_uv(crtc_state, plane_state, plane_id); if (ret) return ret; @@ -5061,8 +5053,7 @@ static int skl_build_plane_wm(struct skl_ddb_allocation *ddb, return 0; } -static int icl_build_plane_wm(struct skl_ddb_allocation *ddb, - struct skl_pipe_wm *pipe_wm, +static int icl_build_plane_wm(struct skl_pipe_wm *pipe_wm, struct intel_crtc_state *crtc_state, const struct intel_plane_state *plane_state) { @@ -5081,17 +5072,17 @@ static int icl_build_plane_wm(struct skl_ddb_allocation *ddb, WARN_ON(!fb->format->is_yuv || fb->format->num_planes == 1); - ret = skl_build_plane_wm_single(ddb, crtc_state, plane_state, + ret = skl_build_plane_wm_single(crtc_state, plane_state, y_plane_id, 0); if (ret) return ret; - ret = skl_build_plane_wm_single(ddb, crtc_state, plane_state, + ret = skl_build_plane_wm_single(crtc_state, plane_state, plane_id, 1); if (ret) return ret; } else if (intel_wm_plane_visible(crtc_state, plane_state)) { - ret = skl_build_plane_wm_single(ddb, crtc_state, plane_state, + ret = skl_build_plane_wm_single(crtc_state, plane_state, plane_id, 0); if (ret) return ret; @@ -5101,7 +5092,6 @@ static int icl_build_plane_wm(struct skl_ddb_allocation *ddb, } static int skl_build_pipe_wm(struct intel_crtc_state *cstate, - struct skl_ddb_allocation *ddb, struct skl_pipe_wm *pipe_wm) { struct drm_i915_private *dev_priv = to_i915(cstate->base.crtc->dev); @@ -5121,10 +5111,10 @@ static int skl_build_pipe_wm(struct intel_crtc_state *cstate, to_intel_plane_state(pstate); if (INTEL_GEN(dev_priv) >= 11) - ret = icl_build_plane_wm(ddb, pipe_wm, + ret = icl_build_plane_wm(pipe_wm, cstate, intel_pstate); else - ret = skl_build_plane_wm(ddb, pipe_wm, + ret = skl_build_plane_wm(pipe_wm, cstate, intel_pstate); if (ret) return ret; @@ -5140,9 +5130,9 @@ static void skl_ddb_entry_write(struct drm_i915_private *dev_priv, const struct skl_ddb_entry *entry) { if (entry->end) - I915_WRITE(reg, (entry->end - 1) << 16 | entry->start); + I915_WRITE_FW(reg, (entry->end - 1) << 16 | entry->start); else - I915_WRITE(reg, 0); + I915_WRITE_FW(reg, 0); } static void skl_write_wm_level(struct drm_i915_private *dev_priv, @@ -5157,19 +5147,22 @@ static void skl_write_wm_level(struct drm_i915_private *dev_priv, val |= level->plane_res_l << PLANE_WM_LINES_SHIFT; } - I915_WRITE(reg, val); + I915_WRITE_FW(reg, val); } -static void skl_write_plane_wm(struct intel_crtc *intel_crtc, - const struct skl_plane_wm *wm, - const struct skl_ddb_allocation *ddb, - enum plane_id plane_id) +void skl_write_plane_wm(struct intel_plane *plane, + const struct intel_crtc_state *crtc_state) { - struct drm_crtc *crtc = &intel_crtc->base; - struct drm_device *dev = crtc->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = to_i915(plane->base.dev); int level, max_level = ilk_wm_max_level(dev_priv); - enum pipe pipe = intel_crtc->pipe; + enum plane_id plane_id = plane->id; + enum pipe pipe = plane->pipe; + const struct skl_plane_wm *wm = + &crtc_state->wm.skl.optimal.planes[plane_id]; + const struct skl_ddb_entry *ddb_y = + &crtc_state->wm.skl.plane_ddb_y[plane_id]; + const struct skl_ddb_entry *ddb_uv = + &crtc_state->wm.skl.plane_ddb_uv[plane_id]; for (level = 0; level <= max_level; level++) { skl_write_wm_level(dev_priv, PLANE_WM(pipe, plane_id, level), @@ -5178,29 +5171,32 @@ static void skl_write_plane_wm(struct intel_crtc *intel_crtc, skl_write_wm_level(dev_priv, PLANE_WM_TRANS(pipe, plane_id), &wm->trans_wm); - if (wm->is_planar && INTEL_GEN(dev_priv) < 11) { - skl_ddb_entry_write(dev_priv, PLANE_BUF_CFG(pipe, plane_id), - &ddb->uv_plane[pipe][plane_id]); + if (INTEL_GEN(dev_priv) >= 11) { skl_ddb_entry_write(dev_priv, - PLANE_NV12_BUF_CFG(pipe, plane_id), - &ddb->plane[pipe][plane_id]); - } else { - skl_ddb_entry_write(dev_priv, PLANE_BUF_CFG(pipe, plane_id), - &ddb->plane[pipe][plane_id]); - if (INTEL_GEN(dev_priv) < 11) - I915_WRITE(PLANE_NV12_BUF_CFG(pipe, plane_id), 0x0); + PLANE_BUF_CFG(pipe, plane_id), ddb_y); + return; } + + if (wm->is_planar) + swap(ddb_y, ddb_uv); + + skl_ddb_entry_write(dev_priv, + PLANE_BUF_CFG(pipe, plane_id), ddb_y); + skl_ddb_entry_write(dev_priv, + PLANE_NV12_BUF_CFG(pipe, plane_id), ddb_uv); } -static void skl_write_cursor_wm(struct intel_crtc *intel_crtc, - const struct skl_plane_wm *wm, - const struct skl_ddb_allocation *ddb) +void skl_write_cursor_wm(struct intel_plane *plane, + const struct intel_crtc_state *crtc_state) { - struct drm_crtc *crtc = &intel_crtc->base; - struct drm_device *dev = crtc->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = to_i915(plane->base.dev); int level, max_level = ilk_wm_max_level(dev_priv); - enum pipe pipe = intel_crtc->pipe; + enum plane_id plane_id = plane->id; + enum pipe pipe = plane->pipe; + const struct skl_plane_wm *wm = + &crtc_state->wm.skl.optimal.planes[plane_id]; + const struct skl_ddb_entry *ddb = + &crtc_state->wm.skl.plane_ddb_y[plane_id]; for (level = 0; level <= max_level; level++) { skl_write_wm_level(dev_priv, CUR_WM(pipe, level), @@ -5208,22 +5204,30 @@ static void skl_write_cursor_wm(struct intel_crtc *intel_crtc, } skl_write_wm_level(dev_priv, CUR_WM_TRANS(pipe), &wm->trans_wm); - skl_ddb_entry_write(dev_priv, CUR_BUF_CFG(pipe), - &ddb->plane[pipe][PLANE_CURSOR]); + skl_ddb_entry_write(dev_priv, CUR_BUF_CFG(pipe), ddb); } bool skl_wm_level_equals(const struct skl_wm_level *l1, const struct skl_wm_level *l2) { - if (l1->plane_en != l2->plane_en) - return false; + return l1->plane_en == l2->plane_en && + l1->plane_res_l == l2->plane_res_l && + l1->plane_res_b == l2->plane_res_b; +} - /* If both planes aren't enabled, the rest shouldn't matter */ - if (!l1->plane_en) - return true; +static bool skl_plane_wm_equals(struct drm_i915_private *dev_priv, + const struct skl_plane_wm *wm1, + const struct skl_plane_wm *wm2) +{ + int level, max_level = ilk_wm_max_level(dev_priv); - return (l1->plane_res_l == l2->plane_res_l && - l1->plane_res_b == l2->plane_res_b); + for (level = 0; level <= max_level; level++) { + if (!skl_wm_level_equals(&wm1->wm[level], &wm2->wm[level]) || + !skl_wm_level_equals(&wm1->uv_wm[level], &wm2->uv_wm[level])) + return false; + } + + return skl_wm_level_equals(&wm1->trans_wm, &wm2->trans_wm); } static inline bool skl_ddb_entries_overlap(const struct skl_ddb_entry *a, @@ -5250,13 +5254,12 @@ bool skl_ddb_allocation_overlaps(const struct skl_ddb_entry *ddb, static int skl_update_pipe_wm(struct drm_crtc_state *cstate, const struct skl_pipe_wm *old_pipe_wm, struct skl_pipe_wm *pipe_wm, /* out */ - struct skl_ddb_allocation *ddb, /* out */ bool *changed /* out */) { struct intel_crtc_state *intel_cstate = to_intel_crtc_state(cstate); int ret; - ret = skl_build_pipe_wm(intel_cstate, ddb, pipe_wm); + ret = skl_build_pipe_wm(intel_cstate, pipe_wm); if (ret) return ret; @@ -5282,42 +5285,29 @@ pipes_modified(struct drm_atomic_state *state) } static int -skl_ddb_add_affected_planes(struct intel_crtc_state *cstate) +skl_ddb_add_affected_planes(const struct intel_crtc_state *old_crtc_state, + struct intel_crtc_state *new_crtc_state) { - struct drm_atomic_state *state = cstate->base.state; - struct drm_device *dev = state->dev; - struct drm_crtc *crtc = cstate->base.crtc; - struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - struct drm_i915_private *dev_priv = to_i915(dev); - struct intel_atomic_state *intel_state = to_intel_atomic_state(state); - struct skl_ddb_allocation *new_ddb = &intel_state->wm_results.ddb; - struct skl_ddb_allocation *cur_ddb = &dev_priv->wm.skl_hw.ddb; - struct drm_plane *plane; - enum pipe pipe = intel_crtc->pipe; + struct intel_atomic_state *state = to_intel_atomic_state(new_crtc_state->base.state); + struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->base.crtc); + struct drm_i915_private *dev_priv = to_i915(crtc->base.dev); + struct intel_plane *plane; - drm_for_each_plane_mask(plane, dev, cstate->base.plane_mask) { - struct drm_plane_state *plane_state; - struct intel_plane *linked; - enum plane_id plane_id = to_intel_plane(plane)->id; + for_each_intel_plane_on_crtc(&dev_priv->drm, crtc, plane) { + struct intel_plane_state *plane_state; + enum plane_id plane_id = plane->id; - if (skl_ddb_entry_equal(&cur_ddb->plane[pipe][plane_id], - &new_ddb->plane[pipe][plane_id]) && - skl_ddb_entry_equal(&cur_ddb->uv_plane[pipe][plane_id], - &new_ddb->uv_plane[pipe][plane_id])) + if (skl_ddb_entry_equal(&old_crtc_state->wm.skl.plane_ddb_y[plane_id], + &new_crtc_state->wm.skl.plane_ddb_y[plane_id]) && + skl_ddb_entry_equal(&old_crtc_state->wm.skl.plane_ddb_uv[plane_id], + &new_crtc_state->wm.skl.plane_ddb_uv[plane_id])) continue; - plane_state = drm_atomic_get_plane_state(state, plane); + plane_state = intel_atomic_get_plane_state(state, plane); if (IS_ERR(plane_state)) return PTR_ERR(plane_state); - /* Make sure linked plane is updated too */ - linked = to_intel_plane_state(plane_state)->linked_plane; - if (!linked) - continue; - - plane_state = drm_atomic_get_plane_state(state, &linked->base); - if (IS_ERR(plane_state)) - return PTR_ERR(plane_state); + new_crtc_state->update_planes |= BIT(plane_id); } return 0; @@ -5329,18 +5319,21 @@ skl_compute_ddb(struct drm_atomic_state *state) const struct drm_i915_private *dev_priv = to_i915(state->dev); struct intel_atomic_state *intel_state = to_intel_atomic_state(state); struct skl_ddb_allocation *ddb = &intel_state->wm_results.ddb; + struct intel_crtc_state *old_crtc_state; + struct intel_crtc_state *new_crtc_state; struct intel_crtc *crtc; - struct intel_crtc_state *cstate; int ret, i; memcpy(ddb, &dev_priv->wm.skl_hw.ddb, sizeof(*ddb)); - for_each_new_intel_crtc_in_state(intel_state, crtc, cstate, i) { - ret = skl_allocate_pipe_ddb(cstate, ddb); + for_each_oldnew_intel_crtc_in_state(intel_state, crtc, old_crtc_state, + new_crtc_state, i) { + ret = skl_allocate_pipe_ddb(new_crtc_state, ddb); if (ret) return ret; - ret = skl_ddb_add_affected_planes(cstate); + ret = skl_ddb_add_affected_planes(old_crtc_state, + new_crtc_state); if (ret) return ret; } @@ -5349,36 +5342,29 @@ skl_compute_ddb(struct drm_atomic_state *state) } static void -skl_print_wm_changes(const struct drm_atomic_state *state) +skl_print_wm_changes(struct intel_atomic_state *state) { - const struct drm_device *dev = state->dev; - const struct drm_i915_private *dev_priv = to_i915(dev); - const struct intel_atomic_state *intel_state = - to_intel_atomic_state(state); - const struct drm_crtc *crtc; - const struct drm_crtc_state *cstate; - const struct intel_plane *intel_plane; - const struct skl_ddb_allocation *old_ddb = &dev_priv->wm.skl_hw.ddb; - const struct skl_ddb_allocation *new_ddb = &intel_state->wm_results.ddb; + struct drm_i915_private *dev_priv = to_i915(state->base.dev); + const struct intel_crtc_state *old_crtc_state; + const struct intel_crtc_state *new_crtc_state; + struct intel_plane *plane; + struct intel_crtc *crtc; int i; - for_each_new_crtc_in_state(state, crtc, cstate, i) { - const struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - enum pipe pipe = intel_crtc->pipe; - - for_each_intel_plane_on_crtc(dev, intel_crtc, intel_plane) { - enum plane_id plane_id = intel_plane->id; + for_each_oldnew_intel_crtc_in_state(state, crtc, old_crtc_state, + new_crtc_state, i) { + for_each_intel_plane_on_crtc(&dev_priv->drm, crtc, plane) { + enum plane_id plane_id = plane->id; const struct skl_ddb_entry *old, *new; - old = &old_ddb->plane[pipe][plane_id]; - new = &new_ddb->plane[pipe][plane_id]; + old = &old_crtc_state->wm.skl.plane_ddb_y[plane_id]; + new = &new_crtc_state->wm.skl.plane_ddb_y[plane_id]; if (skl_ddb_entry_equal(old, new)) continue; DRM_DEBUG_KMS("[PLANE:%d:%s] ddb (%d - %d) -> (%d - %d)\n", - intel_plane->base.base.id, - intel_plane->base.name, + plane->base.base.id, plane->base.name, old->start, old->end, new->start, new->end); } @@ -5475,6 +5461,66 @@ skl_ddb_add_affected_pipes(struct drm_atomic_state *state, bool *changed) return 0; } +/* + * To make sure the cursor watermark registers are always consistent + * with our computed state the following scenario needs special + * treatment: + * + * 1. enable cursor + * 2. move cursor entirely offscreen + * 3. disable cursor + * + * Step 2. does call .disable_plane() but does not zero the watermarks + * (since we consider an offscreen cursor still active for the purposes + * of watermarks). Step 3. would not normally call .disable_plane() + * because the actual plane visibility isn't changing, and we don't + * deallocate the cursor ddb until the pipe gets disabled. So we must + * force step 3. to call .disable_plane() to update the watermark + * registers properly. + * + * Other planes do not suffer from this issues as their watermarks are + * calculated based on the actual plane visibility. The only time this + * can trigger for the other planes is during the initial readout as the + * default value of the watermarks registers is not zero. + */ +static int skl_wm_add_affected_planes(struct intel_atomic_state *state, + struct intel_crtc *crtc) +{ + struct drm_i915_private *dev_priv = to_i915(crtc->base.dev); + const struct intel_crtc_state *old_crtc_state = + intel_atomic_get_old_crtc_state(state, crtc); + struct intel_crtc_state *new_crtc_state = + intel_atomic_get_new_crtc_state(state, crtc); + struct intel_plane *plane; + + for_each_intel_plane_on_crtc(&dev_priv->drm, crtc, plane) { + struct intel_plane_state *plane_state; + enum plane_id plane_id = plane->id; + + /* + * Force a full wm update for every plane on modeset. + * Required because the reset value of the wm registers + * is non-zero, whereas we want all disabled planes to + * have zero watermarks. So if we turn off the relevant + * power well the hardware state will go out of sync + * with the software state. + */ + if (!drm_atomic_crtc_needs_modeset(&new_crtc_state->base) && + skl_plane_wm_equals(dev_priv, + &old_crtc_state->wm.skl.optimal.planes[plane_id], + &new_crtc_state->wm.skl.optimal.planes[plane_id])) + continue; + + plane_state = intel_atomic_get_plane_state(state, plane); + if (IS_ERR(plane_state)) + return PTR_ERR(plane_state); + + new_crtc_state->update_planes |= BIT(plane_id); + } + + return 0; +} + static int skl_compute_wm(struct drm_atomic_state *state) { @@ -5514,8 +5560,12 @@ skl_compute_wm(struct drm_atomic_state *state) &to_intel_crtc_state(crtc->state)->wm.skl.optimal; pipe_wm = &intel_cstate->wm.skl.optimal; - ret = skl_update_pipe_wm(cstate, old_pipe_wm, pipe_wm, - &results->ddb, &changed); + ret = skl_update_pipe_wm(cstate, old_pipe_wm, pipe_wm, &changed); + if (ret) + return ret; + + ret = skl_wm_add_affected_planes(intel_state, + to_intel_crtc(crtc)); if (ret) return ret; @@ -5529,7 +5579,7 @@ skl_compute_wm(struct drm_atomic_state *state) intel_cstate->update_wm_pre = true; } - skl_print_wm_changes(state); + skl_print_wm_changes(intel_state); return 0; } @@ -5540,23 +5590,12 @@ static void skl_atomic_update_crtc_wm(struct intel_atomic_state *state, struct intel_crtc *crtc = to_intel_crtc(cstate->base.crtc); struct drm_i915_private *dev_priv = to_i915(state->base.dev); struct skl_pipe_wm *pipe_wm = &cstate->wm.skl.optimal; - const struct skl_ddb_allocation *ddb = &state->wm_results.ddb; enum pipe pipe = crtc->pipe; - enum plane_id plane_id; if (!(state->wm_results.dirty_pipes & drm_crtc_mask(&crtc->base))) return; I915_WRITE(PIPE_WM_LINETIME(pipe), pipe_wm->linetime); - - for_each_plane_id_on_crtc(crtc, plane_id) { - if (plane_id != PLANE_CURSOR) - skl_write_plane_wm(crtc, &pipe_wm->planes[plane_id], - ddb, plane_id); - else - skl_write_cursor_wm(crtc, &pipe_wm->planes[plane_id], - ddb); - } } static void skl_initial_wm(struct intel_atomic_state *state, @@ -5566,8 +5605,6 @@ static void skl_initial_wm(struct intel_atomic_state *state, struct drm_device *dev = intel_crtc->base.dev; struct drm_i915_private *dev_priv = to_i915(dev); struct skl_ddb_values *results = &state->wm_results; - struct skl_ddb_values *hw_vals = &dev_priv->wm.skl_hw; - enum pipe pipe = intel_crtc->pipe; if ((results->dirty_pipes & drm_crtc_mask(&intel_crtc->base)) == 0) return; @@ -5577,11 +5614,6 @@ static void skl_initial_wm(struct intel_atomic_state *state, if (cstate->base.active_changed) skl_atomic_update_crtc_wm(state, cstate); - memcpy(hw_vals->ddb.uv_plane[pipe], results->ddb.uv_plane[pipe], - sizeof(hw_vals->ddb.uv_plane[pipe])); - memcpy(hw_vals->ddb.plane[pipe], results->ddb.plane[pipe], - sizeof(hw_vals->ddb.plane[pipe])); - mutex_unlock(&dev_priv->wm.wm_mutex); } @@ -5732,13 +5764,6 @@ void skl_wm_get_hw_state(struct drm_device *dev) if (dev_priv->active_crtcs) { /* Fully recompute DDB on first atomic commit */ dev_priv->wm.distrust_bios_wm = true; - } else { - /* - * Easy/common case; just sanitize DDB now if everything off - * Keep dbuf slice info intact - */ - memset(ddb->plane, 0, sizeof(ddb->plane)); - memset(ddb->uv_plane, 0, sizeof(ddb->uv_plane)); } } diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c index e5da2fe3bdae..32df604e90b6 100644 --- a/drivers/gpu/drm/i915/intel_sprite.c +++ b/drivers/gpu/drm/i915/intel_sprite.c @@ -542,6 +542,8 @@ skl_program_plane(struct intel_plane *plane, if (fb->format->is_yuv && icl_is_hdr_plane(plane)) icl_program_input_csc_coeff(crtc_state, plane_state); + skl_write_plane_wm(plane, crtc_state); + I915_WRITE_FW(PLANE_KEYVAL(pipe, plane_id), key->min_value); I915_WRITE_FW(PLANE_KEYMSK(pipe, plane_id), keymsk); I915_WRITE_FW(PLANE_KEYMAX(pipe, plane_id), keymax); @@ -604,6 +606,8 @@ skl_disable_plane(struct intel_plane *plane, spin_lock_irqsave(&dev_priv->uncore.lock, irqflags); + skl_write_plane_wm(plane, crtc_state); + I915_WRITE_FW(PLANE_CTL(pipe, plane_id), 0); I915_WRITE_FW(PLANE_SURF(pipe, plane_id), 0); From 5f2e511205bbc40b6f58f7d2125c10bd39007748 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Wed, 14 Nov 2018 23:07:27 +0200 Subject: [PATCH 29/98] drm/i915: Commit skl+ planes in an order that avoids ddb overlaps MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit skl+ can go belly up if there are overlapping ddb allocations between planes. If we could absolutely guarantee that we can perform the atomic update within a single frame we shouldn't have to worry about this. But we can't rely on that so let's steal the ddb overlap check trick from skl_update_crtcs() and apply it to the plane updates. Since each step of the sequence is free from ddb overlaps we don't have to worry about a vblank sneaking up on us in the middle of the sequence. The partial state that gets latched by the hardware will be safe. And unlike skl_update_crtcs() we don't have to intoduce any extra vblank waits on account of only having to worry about a single pipe. v2: Fix typo in commit msg (Matt) Signed-off-by: Ville Syrjälä Link: https://patchwork.freedesktop.org/patch/msgid/20181114210729.16185-12-ville.syrjala@linux.intel.com Reviewed-by: Matt Roper Reviewed-by: Maarten Lankhorst --- drivers/gpu/drm/i915/intel_atomic_plane.c | 96 ++++++++++++++++++++--- drivers/gpu/drm/i915/intel_display.c | 7 +- drivers/gpu/drm/i915/intel_drv.h | 8 +- 3 files changed, 93 insertions(+), 18 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_atomic_plane.c b/drivers/gpu/drm/i915/intel_atomic_plane.c index ad3d135bdc15..0a73e6e65c20 100644 --- a/drivers/gpu/drm/i915/intel_atomic_plane.c +++ b/drivers/gpu/drm/i915/intel_atomic_plane.c @@ -171,24 +171,75 @@ static int intel_plane_atomic_check(struct drm_plane *plane, to_intel_plane_state(new_plane_state)); } -void intel_update_planes_on_crtc(struct intel_atomic_state *old_state, - struct intel_crtc *crtc, - struct intel_crtc_state *old_crtc_state, - struct intel_crtc_state *new_crtc_state) +static struct intel_plane * +skl_next_plane_to_commit(struct intel_atomic_state *state, + struct intel_crtc *crtc, + struct skl_ddb_entry entries_y[I915_MAX_PLANES], + struct skl_ddb_entry entries_uv[I915_MAX_PLANES], + unsigned int *update_mask) { - u32 update_mask = new_crtc_state->update_planes; - struct intel_plane_state *new_plane_state; + struct intel_crtc_state *crtc_state = + intel_atomic_get_new_crtc_state(state, crtc); + struct intel_plane_state *plane_state; struct intel_plane *plane; int i; - for_each_new_intel_plane_in_state(old_state, plane, new_plane_state, i) { + if (*update_mask == 0) + return NULL; + + for_each_new_intel_plane_in_state(state, plane, plane_state, i) { + enum plane_id plane_id = plane->id; + if (crtc->pipe != plane->pipe || - !(update_mask & BIT(plane->id))) + !(*update_mask & BIT(plane_id))) continue; + if (skl_ddb_allocation_overlaps(&crtc_state->wm.skl.plane_ddb_y[plane_id], + entries_y, + I915_MAX_PLANES, plane_id) || + skl_ddb_allocation_overlaps(&crtc_state->wm.skl.plane_ddb_uv[plane_id], + entries_uv, + I915_MAX_PLANES, plane_id)) + continue; + + *update_mask &= ~BIT(plane_id); + entries_y[plane_id] = crtc_state->wm.skl.plane_ddb_y[plane_id]; + entries_uv[plane_id] = crtc_state->wm.skl.plane_ddb_uv[plane_id]; + + return plane; + } + + /* should never happen */ + WARN_ON(1); + + return NULL; +} + +void skl_update_planes_on_crtc(struct intel_atomic_state *state, + struct intel_crtc *crtc) +{ + struct intel_crtc_state *old_crtc_state = + intel_atomic_get_old_crtc_state(state, crtc); + struct intel_crtc_state *new_crtc_state = + intel_atomic_get_new_crtc_state(state, crtc); + struct skl_ddb_entry entries_y[I915_MAX_PLANES]; + struct skl_ddb_entry entries_uv[I915_MAX_PLANES]; + u32 update_mask = new_crtc_state->update_planes; + struct intel_plane *plane; + + memcpy(entries_y, old_crtc_state->wm.skl.plane_ddb_y, + sizeof(old_crtc_state->wm.skl.plane_ddb_y)); + memcpy(entries_uv, old_crtc_state->wm.skl.plane_ddb_uv, + sizeof(old_crtc_state->wm.skl.plane_ddb_uv)); + + while ((plane = skl_next_plane_to_commit(state, crtc, + entries_y, entries_uv, + &update_mask))) { + struct intel_plane_state *new_plane_state = + intel_atomic_get_new_plane_state(state, plane); + if (new_plane_state->base.visible) { trace_intel_update_plane(&plane->base, crtc); - plane->update_plane(plane, new_crtc_state, new_plane_state); } else if (new_plane_state->slave) { struct intel_plane *master = @@ -204,14 +255,37 @@ void intel_update_planes_on_crtc(struct intel_atomic_state *old_state, * plane_state. */ new_plane_state = - intel_atomic_get_new_plane_state(old_state, master); + intel_atomic_get_new_plane_state(state, master); trace_intel_update_plane(&plane->base, crtc); - plane->update_slave(plane, new_crtc_state, new_plane_state); } else { trace_intel_disable_plane(&plane->base, crtc); + plane->disable_plane(plane, new_crtc_state); + } + } +} +void i9xx_update_planes_on_crtc(struct intel_atomic_state *state, + struct intel_crtc *crtc) +{ + struct intel_crtc_state *new_crtc_state = + intel_atomic_get_new_crtc_state(state, crtc); + u32 update_mask = new_crtc_state->update_planes; + struct intel_plane_state *new_plane_state; + struct intel_plane *plane; + int i; + + for_each_new_intel_plane_in_state(state, plane, new_plane_state, i) { + if (crtc->pipe != plane->pipe || + !(update_mask & BIT(plane->id))) + continue; + + if (new_plane_state->base.visible) { + trace_intel_update_plane(&plane->base, crtc); + plane->update_plane(plane, new_crtc_state, new_plane_state); + } else { + trace_intel_disable_plane(&plane->base, crtc); plane->disable_plane(plane, new_crtc_state); } } diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 6c5d71b44740..d07fa4456150 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -12692,7 +12692,6 @@ static void intel_update_crtc(struct drm_crtc *crtc, struct drm_device *dev = crtc->dev; struct drm_i915_private *dev_priv = to_i915(dev); struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - struct intel_crtc_state *old_intel_cstate = to_intel_crtc_state(old_crtc_state); struct intel_crtc_state *pipe_config = to_intel_crtc_state(new_crtc_state); bool modeset = needs_modeset(new_crtc_state); struct intel_plane_state *new_plane_state = @@ -12715,8 +12714,10 @@ static void intel_update_crtc(struct drm_crtc *crtc, intel_begin_crtc_commit(crtc, old_crtc_state); - intel_update_planes_on_crtc(to_intel_atomic_state(state), intel_crtc, - old_intel_cstate, pipe_config); + if (INTEL_GEN(dev_priv) >= 9) + skl_update_planes_on_crtc(to_intel_atomic_state(state), intel_crtc); + else + i9xx_update_planes_on_crtc(to_intel_atomic_state(state), intel_crtc); intel_finish_crtc_commit(crtc, old_crtc_state); } diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index 1a3a396862d1..40edb21087a7 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -2300,10 +2300,10 @@ struct drm_plane_state *intel_plane_duplicate_state(struct drm_plane *plane); void intel_plane_destroy_state(struct drm_plane *plane, struct drm_plane_state *state); extern const struct drm_plane_helper_funcs intel_plane_helper_funcs; -void intel_update_planes_on_crtc(struct intel_atomic_state *old_state, - struct intel_crtc *crtc, - struct intel_crtc_state *old_crtc_state, - struct intel_crtc_state *new_crtc_state); +void skl_update_planes_on_crtc(struct intel_atomic_state *state, + struct intel_crtc *crtc); +void i9xx_update_planes_on_crtc(struct intel_atomic_state *state, + struct intel_crtc *crtc); int intel_plane_atomic_check_with_state(const struct intel_crtc_state *old_crtc_state, struct intel_crtc_state *crtc_state, const struct intel_plane_state *old_plane_state, From 45bee430b8438a26e0da7c664605ceea97823ec8 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Wed, 14 Nov 2018 23:07:28 +0200 Subject: [PATCH 30/98] drm/i915: Rename the confusing 'plane_id' to 'color_plane' MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit A variable whose name is 'plane_id' is expected to be of the enum plane_id type. In this case we have a raw int, which turns out to refer to the plane of the framebuffer. Rename the variable to 'color_plane' in line with the trend started earlier. Signed-off-by: Ville Syrjälä Link: https://patchwork.freedesktop.org/patch/msgid/20181114210729.16185-13-ville.syrjala@linux.intel.com Reviewed-by: Matt Roper Reviewed-by: Maarten Lankhorst --- drivers/gpu/drm/i915/intel_pm.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index b1f30d56b747..a26b4eddda25 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -4618,7 +4618,7 @@ skl_adjusted_plane_pixel_rate(const struct intel_crtc_state *cstate, static int skl_compute_plane_wm_params(const struct intel_crtc_state *cstate, const struct intel_plane_state *intel_pstate, - struct skl_wm_params *wp, int plane_id) + struct skl_wm_params *wp, int color_plane) { struct intel_plane *plane = to_intel_plane(intel_pstate->base.plane); struct drm_i915_private *dev_priv = to_i915(plane->base.dev); @@ -4630,7 +4630,7 @@ skl_compute_plane_wm_params(const struct intel_crtc_state *cstate, bool apply_memory_bw_wa = skl_needs_memory_bw_wa(state); /* only NV12 format has two planes */ - if (plane_id == 1 && fb->format->format != DRM_FORMAT_NV12) { + if (color_plane == 1 && fb->format->format != DRM_FORMAT_NV12) { DRM_DEBUG_KMS("Non NV12 format have single plane\n"); return -EINVAL; } @@ -4655,10 +4655,10 @@ skl_compute_plane_wm_params(const struct intel_crtc_state *cstate, wp->width = drm_rect_width(&intel_pstate->base.src) >> 16; } - if (plane_id == 1 && wp->is_planar) + if (color_plane == 1 && wp->is_planar) wp->width /= 2; - wp->cpp = fb->format->cpp[plane_id]; + wp->cpp = fb->format->cpp[color_plane]; wp->plane_pixel_rate = skl_adjusted_plane_pixel_rate(cstate, intel_pstate); From 1fdee7582cce4c9d56b24e6caf1a87cd47a11825 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Wed, 14 Nov 2018 23:07:29 +0200 Subject: [PATCH 31/98] drm/i915: Pass the plane to icl_program_input_csc_coeff() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit On icl+ the plane state that gets passed to update_slave() is not the plane state of the plane we're programming. With NV12 the plane state would be coming from the master (UV) plane whereas the plane we're programming is the slave (Y) plane. For that reason we need to explicitly pass around the slave plane (or we'd have to otherwise deduce it by checking whether we were called via .update_plane() or .update_slave()). In the case of icl_program_input_csc_coeff() it's actually OK to assume that we are always the master plane because the input CSC only exists on HDR planes which can never be a slave plane. But for consistency let's pass in the plane explicitly anyway. While at it drop the "_coeff" from the function name since it's kinda redundant, and this makes the name a bit shorter :) Cc: Uma Shankar Cc: Maarten Lankhorst Signed-off-by: Ville Syrjälä Link: https://patchwork.freedesktop.org/patch/msgid/20181114210729.16185-14-ville.syrjala@linux.intel.com Reviewed-by: Uma Shankar Reviewed-by: Maarten Lankhorst Reviewed-by: Matt Roper --- drivers/gpu/drm/i915/intel_sprite.c | 14 ++++++-------- 1 file changed, 6 insertions(+), 8 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c index 32df604e90b6..d2e003d8f3db 100644 --- a/drivers/gpu/drm/i915/intel_sprite.c +++ b/drivers/gpu/drm/i915/intel_sprite.c @@ -373,14 +373,12 @@ skl_program_scaler(struct intel_plane *plane, #define BOFF(x) (((x) & 0xffff) << 16) static void -icl_program_input_csc_coeff(const struct intel_crtc_state *crtc_state, - const struct intel_plane_state *plane_state) +icl_program_input_csc(struct intel_plane *plane, + const struct intel_crtc_state *crtc_state, + const struct intel_plane_state *plane_state) { - struct drm_i915_private *dev_priv = - to_i915(plane_state->base.plane->dev); - struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc); - enum pipe pipe = crtc->pipe; - struct intel_plane *plane = to_intel_plane(plane_state->base.plane); + struct drm_i915_private *dev_priv = to_i915(plane->base.dev); + enum pipe pipe = plane->pipe; enum plane_id plane_id = plane->id; static const u16 input_csc_matrix[][9] = { @@ -540,7 +538,7 @@ skl_program_plane(struct intel_plane *plane, plane_state->color_ctl); if (fb->format->is_yuv && icl_is_hdr_plane(plane)) - icl_program_input_csc_coeff(crtc_state, plane_state); + icl_program_input_csc(plane, crtc_state, plane_state); skl_write_plane_wm(plane, crtc_state); From 70bbe53c6eacdf01485848b7568609e33372cad0 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Tue, 23 Oct 2018 19:02:01 +0300 Subject: [PATCH 32/98] drm/i915: Make sure fb gtt offsets stay within 32bits MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Let's try to make sure the fb offset computations never hit an integer overflow by making sure the entire fb stays below 32bits. framebuffer_check() in the core already does the same check, but as it doesn't know about tiling some things can slip through. Repeat the check in the driver with tiling taken into account. v2: Use add_overflows() after massaging it to work for me (Chris) v3: Call it add_overflow_t() to match min_t() & co. (Chris) Cc: Chris Wilson Signed-off-by: Ville Syrjälä Link: https://patchwork.freedesktop.org/patch/msgid/20181023160201.9840-1-ville.syrjala@linux.intel.com Reviewed-by: Chris Wilson --- drivers/gpu/drm/i915/i915_utils.h | 11 +++++++---- drivers/gpu/drm/i915/intel_display.c | 18 +++++++++++++++++- 2 files changed, 24 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_utils.h b/drivers/gpu/drm/i915/i915_utils.h index 5858a43e19da..9726df37c4c4 100644 --- a/drivers/gpu/drm/i915/i915_utils.h +++ b/drivers/gpu/drm/i915/i915_utils.h @@ -44,16 +44,19 @@ __stringify(x), (long)(x)) #if defined(GCC_VERSION) && GCC_VERSION >= 70000 -#define add_overflows(A, B) \ - __builtin_add_overflow_p((A), (B), (typeof((A) + (B)))0) +#define add_overflows_t(T, A, B) \ + __builtin_add_overflow_p((A), (B), (T)0) #else -#define add_overflows(A, B) ({ \ +#define add_overflows_t(T, A, B) ({ \ typeof(A) a = (A); \ typeof(B) b = (B); \ - a + b < a; \ + (T)(a + b) < a; \ }) #endif +#define add_overflows(A, B) \ + add_overflows_t(typeof((A) + (B)), (A), (B)) + #define range_overflows(start, size, max) ({ \ typeof(start) start__ = (start); \ typeof(size) size__ = (size); \ diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index d07fa4456150..14cca56f2743 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -2341,10 +2341,26 @@ static int intel_fb_offset_to_xy(int *x, int *y, int color_plane) { struct drm_i915_private *dev_priv = to_i915(fb->dev); + unsigned int height; if (fb->modifier != DRM_FORMAT_MOD_LINEAR && - fb->offsets[color_plane] % intel_tile_size(dev_priv)) + fb->offsets[color_plane] % intel_tile_size(dev_priv)) { + DRM_DEBUG_KMS("Misaligned offset 0x%08x for color plane %d\n", + fb->offsets[color_plane], color_plane); return -EINVAL; + } + + height = drm_framebuffer_plane_height(fb->height, fb, color_plane); + height = ALIGN(height, intel_tile_height(fb, color_plane)); + + /* Catch potential overflows early */ + if (add_overflows_t(u32, mul_u32_u32(height, fb->pitches[color_plane]), + fb->offsets[color_plane])) { + DRM_DEBUG_KMS("Bad offset 0x%08x or pitch %d for color plane %d\n", + fb->offsets[color_plane], fb->pitches[color_plane], + color_plane); + return -ERANGE; + } *x = 0; *y = 0; From b3cf5c06ca500150fe17b72f411d0fb620b2f864 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Tue, 25 Sep 2018 22:37:08 +0300 Subject: [PATCH 33/98] drm/i915: Decouple SKL stride units from intel_fb_stride_alignment() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit In the future framebuffer stride alignment requirements won't exactly match the units in which skl+ plane stride is specified. So extract the code for the skl+ stuff into a separate helper. Signed-off-by: Ville Syrjälä Link: https://patchwork.freedesktop.org/patch/msgid/20180925193714.25280-3-ville.syrjala@linux.intel.com Reviewed-by: Chris Wilson --- drivers/gpu/drm/i915/intel_display.c | 28 +++++++++++++++++----------- 1 file changed, 17 insertions(+), 11 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 14cca56f2743..6fce44b56be2 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -3475,6 +3475,21 @@ static void skl_detach_scalers(const struct intel_crtc_state *crtc_state) } } +static unsigned int skl_plane_stride_mult(const struct drm_framebuffer *fb, + int color_plane, unsigned int rotation) +{ + /* + * The stride is either expressed as a multiple of 64 bytes chunks for + * linear buffers or in number of tiles for tiled buffers. + */ + if (fb->modifier == DRM_FORMAT_MOD_LINEAR) + return 64; + else if (drm_rotation_90_or_270(rotation)) + return intel_tile_height(fb, color_plane); + else + return intel_tile_width_bytes(fb, color_plane); +} + u32 skl_plane_stride(const struct intel_plane_state *plane_state, int color_plane) { @@ -3485,16 +3500,7 @@ u32 skl_plane_stride(const struct intel_plane_state *plane_state, if (color_plane >= fb->format->num_planes) return 0; - /* - * The stride is either expressed as a multiple of 64 bytes chunks for - * linear buffers or in number of tiles for tiled buffers. - */ - if (drm_rotation_90_or_270(rotation)) - stride /= intel_tile_height(fb, color_plane); - else - stride /= intel_fb_stride_alignment(fb, color_plane); - - return stride; + return stride / skl_plane_stride_mult(fb, color_plane, rotation); } static u32 skl_plane_ctl_format(uint32_t pixel_format) @@ -8967,7 +8973,7 @@ skylake_get_initial_plane_config(struct intel_crtc *crtc, fb->width = ((val >> 0) & 0x1fff) + 1; val = I915_READ(PLANE_STRIDE(pipe, plane_id)); - stride_mult = intel_fb_stride_alignment(fb, 0); + stride_mult = skl_plane_stride_mult(fb, 0, DRM_MODE_ROTATE_0); fb->pitches[0] = (val & 0x3ff) * stride_mult; aligned_height = intel_fb_align_height(fb, 0, fb->height); From 7b610f1fbed2ab3b82a6064a2b5dedf061f31fa6 Mon Sep 17 00:00:00 2001 From: Manasi Navare Date: Wed, 28 Nov 2018 12:26:12 -0800 Subject: [PATCH 34/98] drm/i915/dp: Add DSC params and DSC config to intel_crtc_state Basic DSC parameters and DSC configuration data needs to be computed for each of the requested mode during atomic check. This is required since for certain modes, valid DSC parameters and config data might not be computed in which case compression cannot be enabled for that mode. For that reason we need to add these params and config structure to the intel_crtc_state so that if valid this state information can directly be used while enabling DSC in atomic commit. v2: * Rebase on drm-tip (Manasi) Cc: Gaurav K Singh Cc: Jani Nikula Cc: Ville Syrjala Cc: Anusha Srivatsa Signed-off-by: Manasi Navare Reviewed-by: Anusha Srivatsa Link: https://patchwork.freedesktop.org/patch/msgid/20181128202628.20238-1-manasi.d.navare@intel.com --- drivers/gpu/drm/i915/i915_drv.h | 1 + drivers/gpu/drm/i915/intel_drv.h | 9 +++++++++ 2 files changed, 10 insertions(+) diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 645c2bbdcdfa..42704e731364 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -53,6 +53,7 @@ #include #include #include +#include #include "i915_fixed.h" #include "i915_params.h" diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index 40edb21087a7..5577f0f31b99 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -942,6 +942,15 @@ struct intel_crtc_state { /* Output down scaling is done in LSPCON device */ bool lspcon_downsampling; + + /* Display Stream compression state */ + struct { + bool compression_enable; + bool dsc_split; + u16 compressed_bpp; + u8 slice_count; + } dsc_params; + struct drm_dsc_config dp_dsc_cfg; }; struct intel_crtc { From a4a157777c807d5c0ead1b747dd9b07a07864498 Mon Sep 17 00:00:00 2001 From: Manasi Navare Date: Wed, 28 Nov 2018 13:36:21 -0800 Subject: [PATCH 35/98] drm/i915/dp: Compute DSC pipe config in atomic check DSC params like the enable, compressed bpp, slice count and dsc_split are added to the intel_crtc_state. These parameters are set based on the requested mode and available link parameters during the pipe configuration in atomic check phase. These values are then later used to populate the remaining DSC and RC parameters before enbaling DSC in atomic commit. v15: * Rebase over drm-tip v14: Remove leftovers, use dsc_bpc, refine dsc_compute_config (Ville) v13: * Compute DSC bpc only when DSC is req to be enabled (Ville) v12: * Override bpp with dsc dpcd color depth (Manasi) v11: * Const crtc_state, reject DSC on DP without FEC (Ville) * Dont set dsc_split to false (Ville) v10: * Add a helper for dp_dsc support (Ville) * Set pipe_config to max bpp, link params for DSC for now (Ville) * Compute bpp - use dp dsc support helper (Ville) v9: * Rebase on top of drm-tip that now uses fast_narrow config for edp (Manasi) v8: * Check for DSC bpc not 0 (manasi) v7: * Fix indentation in compute_m_n (Manasi) v6 (From Gaurav): * Remove function call of intel_dp_compute_dsc_params() and invoke intel_dp_compute_dsc_params() in the patch where it is defined to fix compilation warning (Gaurav) v5: Add drm_dsc_cfg in intel_crtc_state (Manasi) v4: * Rebase on refactoring of intel_dp_compute_config on tip (Manasi) * Add a comment why we need to check PSR while enabling DSC (Gaurav) v3: * Check PPR > max_cdclock to use 2 VDSC instances (Ville) v2: * Add if-else for eDP/DP (Gaurav) Cc: Jani Nikula Cc: Ville Syrjala Cc: Anusha Srivatsa Cc: Gaurav K Singh Signed-off-by: Manasi Navare Reviewed-by: Anusha Srivatsa Reviewed-by: Ville Syrjala Acked-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/20181128213621.21391-1-manasi.d.navare@intel.com --- drivers/gpu/drm/i915/intel_display.c | 2 +- drivers/gpu/drm/i915/intel_display.h | 3 +- drivers/gpu/drm/i915/intel_dp.c | 192 ++++++++++++++++++++++++--- 3 files changed, 172 insertions(+), 25 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 6fce44b56be2..f3ba04562cbd 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -6758,7 +6758,7 @@ static void compute_m_n(unsigned int m, unsigned int n, } void -intel_link_compute_m_n(int bits_per_pixel, int nlanes, +intel_link_compute_m_n(u16 bits_per_pixel, int nlanes, int pixel_clock, int link_clock, struct intel_link_m_n *m_n, bool constant_n) diff --git a/drivers/gpu/drm/i915/intel_display.h b/drivers/gpu/drm/i915/intel_display.h index a7ceb8f904f7..3ec704c93d6e 100644 --- a/drivers/gpu/drm/i915/intel_display.h +++ b/drivers/gpu/drm/i915/intel_display.h @@ -440,10 +440,9 @@ struct intel_link_m_n { (__i)++) \ for_each_if(crtc) -void intel_link_compute_m_n(int bpp, int nlanes, +void intel_link_compute_m_n(u16 bpp, int nlanes, int pixel_clock, int link_clock, struct intel_link_m_n *m_n, bool constant_n); - bool is_ccs_modifier(u64 modifier); #endif diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 70ae3d57316b..a2780733768a 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -47,6 +47,8 @@ /* DP DSC small joiner has 2 FIFOs each of 640 x 6 bytes */ #define DP_DSC_MAX_SMALL_JOINER_RAM_BUFFER 61440 +#define DP_DSC_MIN_SUPPORTED_BPC 8 +#define DP_DSC_MAX_SUPPORTED_BPC 10 /* DP DSC throughput values used for slice count calculations KPixels/s */ #define DP_DSC_PEAK_PIXEL_RATE 2720000 @@ -1708,6 +1710,26 @@ struct link_config_limits { int min_bpp, max_bpp; }; +static bool intel_dp_source_supports_dsc(struct intel_dp *intel_dp, + const struct intel_crtc_state *pipe_config) +{ + struct drm_i915_private *dev_priv = dp_to_i915(intel_dp); + + /* FIXME: FEC needed for external DP until then reject DSC on DP */ + if (!intel_dp_is_edp(intel_dp)) + return false; + + return INTEL_GEN(dev_priv) >= 10 && + pipe_config->cpu_transcoder != TRANSCODER_A; +} + +static bool intel_dp_supports_dsc(struct intel_dp *intel_dp, + const struct intel_crtc_state *pipe_config) +{ + return intel_dp_source_supports_dsc(intel_dp, pipe_config) && + drm_dp_sink_supports_dsc(intel_dp->dsc_dpcd); +} + static int intel_dp_compute_bpp(struct intel_dp *intel_dp, struct intel_crtc_state *pipe_config) { @@ -1842,14 +1864,115 @@ intel_dp_compute_link_config_fast(struct intel_dp *intel_dp, return false; } +static int intel_dp_dsc_compute_bpp(struct intel_dp *intel_dp, u8 dsc_max_bpc) +{ + int i, num_bpc; + u8 dsc_bpc[3] = {0}; + + num_bpc = drm_dp_dsc_sink_supported_input_bpcs(intel_dp->dsc_dpcd, + dsc_bpc); + for (i = 0; i < num_bpc; i++) { + if (dsc_max_bpc >= dsc_bpc[i]) + return dsc_bpc[i] * 3; + } + + return 0; +} + +static bool intel_dp_dsc_compute_config(struct intel_dp *intel_dp, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state, + struct link_config_limits *limits) +{ + struct intel_digital_port *dig_port = dp_to_dig_port(intel_dp); + struct drm_i915_private *dev_priv = to_i915(dig_port->base.base.dev); + struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode; + u8 dsc_max_bpc; + int pipe_bpp; + + if (!intel_dp_supports_dsc(intel_dp, pipe_config)) + return false; + + dsc_max_bpc = min_t(u8, DP_DSC_MAX_SUPPORTED_BPC, + conn_state->max_requested_bpc); + + pipe_bpp = intel_dp_dsc_compute_bpp(intel_dp, dsc_max_bpc); + if (pipe_bpp < DP_DSC_MIN_SUPPORTED_BPC * 3) { + DRM_DEBUG_KMS("No DSC support for less than 8bpc\n"); + return false; + } + + /* + * For now enable DSC for max bpp, max link rate, max lane count. + * Optimize this later for the minimum possible link rate/lane count + * with DSC enabled for the requested mode. + */ + pipe_config->pipe_bpp = pipe_bpp; + pipe_config->port_clock = intel_dp->common_rates[limits->max_clock]; + pipe_config->lane_count = limits->max_lane_count; + + if (intel_dp_is_edp(intel_dp)) { + pipe_config->dsc_params.compressed_bpp = + min_t(u16, drm_edp_dsc_sink_output_bpp(intel_dp->dsc_dpcd) >> 4, + pipe_config->pipe_bpp); + pipe_config->dsc_params.slice_count = + drm_dp_dsc_sink_max_slice_count(intel_dp->dsc_dpcd, + true); + } else { + u16 dsc_max_output_bpp; + u8 dsc_dp_slice_count; + + dsc_max_output_bpp = + intel_dp_dsc_get_output_bpp(pipe_config->port_clock, + pipe_config->lane_count, + adjusted_mode->crtc_clock, + adjusted_mode->crtc_hdisplay); + dsc_dp_slice_count = + intel_dp_dsc_get_slice_count(intel_dp, + adjusted_mode->crtc_clock, + adjusted_mode->crtc_hdisplay); + if (!dsc_max_output_bpp || !dsc_dp_slice_count) { + DRM_DEBUG_KMS("Compressed BPP/Slice Count not supported\n"); + return false; + } + pipe_config->dsc_params.compressed_bpp = min_t(u16, + dsc_max_output_bpp >> 4, + pipe_config->pipe_bpp); + pipe_config->dsc_params.slice_count = dsc_dp_slice_count; + } + /* + * VDSC engine operates at 1 Pixel per clock, so if peak pixel rate + * is greater than the maximum Cdclock and if slice count is even + * then we need to use 2 VDSC instances. + */ + if (adjusted_mode->crtc_clock > dev_priv->max_cdclk_freq) { + if (pipe_config->dsc_params.slice_count > 1) { + pipe_config->dsc_params.dsc_split = true; + } else { + DRM_DEBUG_KMS("Cannot split stream to use 2 VDSC instances\n"); + return false; + } + } + pipe_config->dsc_params.compression_enable = true; + DRM_DEBUG_KMS("DP DSC computed with Input Bpp = %d " + "Compressed Bpp = %d Slice Count = %d\n", + pipe_config->pipe_bpp, + pipe_config->dsc_params.compressed_bpp, + pipe_config->dsc_params.slice_count); + + return true; +} + static bool intel_dp_compute_link_config(struct intel_encoder *encoder, - struct intel_crtc_state *pipe_config) + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode; struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base); struct link_config_limits limits; int common_len; + bool ret; common_len = intel_dp_common_len_rate_limit(intel_dp, intel_dp->max_link_rate); @@ -1888,7 +2011,7 @@ intel_dp_compute_link_config(struct intel_encoder *encoder, intel_dp->common_rates[limits.max_clock], limits.max_bpp, adjusted_mode->crtc_clock); - if (intel_dp_is_edp(intel_dp)) { + if (intel_dp_is_edp(intel_dp)) /* * Optimize for fast and narrow. eDP 1.3 section 3.3 and eDP 1.4 * section A.1: "It is recommended that the minimum number of @@ -1898,26 +2021,42 @@ intel_dp_compute_link_config(struct intel_encoder *encoder, * Note that we use the max clock and lane count for eDP 1.3 and * earlier, and fast vs. wide is irrelevant. */ - if (!intel_dp_compute_link_config_fast(intel_dp, pipe_config, - &limits)) - return false; - } else { + ret = intel_dp_compute_link_config_fast(intel_dp, pipe_config, + &limits); + else /* Optimize for slow and wide. */ - if (!intel_dp_compute_link_config_wide(intel_dp, pipe_config, - &limits)) + ret = intel_dp_compute_link_config_wide(intel_dp, pipe_config, + &limits); + + /* enable compression if the mode doesn't fit available BW */ + if (!ret) { + if (!intel_dp_dsc_compute_config(intel_dp, pipe_config, + conn_state, &limits)) return false; } - DRM_DEBUG_KMS("DP lane count %d clock %d bpp %d\n", - pipe_config->lane_count, pipe_config->port_clock, - pipe_config->pipe_bpp); + if (pipe_config->dsc_params.compression_enable) { + DRM_DEBUG_KMS("DP lane count %d clock %d Input bpp %d Compressed bpp %d\n", + pipe_config->lane_count, pipe_config->port_clock, + pipe_config->pipe_bpp, + pipe_config->dsc_params.compressed_bpp); - DRM_DEBUG_KMS("DP link rate required %i available %i\n", - intel_dp_link_required(adjusted_mode->crtc_clock, - pipe_config->pipe_bpp), - intel_dp_max_data_rate(pipe_config->port_clock, - pipe_config->lane_count)); + DRM_DEBUG_KMS("DP link rate required %i available %i\n", + intel_dp_link_required(adjusted_mode->crtc_clock, + pipe_config->dsc_params.compressed_bpp), + intel_dp_max_data_rate(pipe_config->port_clock, + pipe_config->lane_count)); + } else { + DRM_DEBUG_KMS("DP lane count %d clock %d bpp %d\n", + pipe_config->lane_count, pipe_config->port_clock, + pipe_config->pipe_bpp); + DRM_DEBUG_KMS("DP link rate required %i available %i\n", + intel_dp_link_required(adjusted_mode->crtc_clock, + pipe_config->pipe_bpp), + intel_dp_max_data_rate(pipe_config->port_clock, + pipe_config->lane_count)); + } return true; } @@ -1983,7 +2122,7 @@ intel_dp_compute_config(struct intel_encoder *encoder, if (adjusted_mode->flags & DRM_MODE_FLAG_DBLCLK) return false; - if (!intel_dp_compute_link_config(encoder, pipe_config)) + if (!intel_dp_compute_link_config(encoder, pipe_config, conn_state)) return false; if (intel_conn_state->broadcast_rgb == INTEL_BROADCAST_RGB_AUTO) { @@ -2001,11 +2140,20 @@ intel_dp_compute_config(struct intel_encoder *encoder, intel_conn_state->broadcast_rgb == INTEL_BROADCAST_RGB_LIMITED; } - intel_link_compute_m_n(pipe_config->pipe_bpp, pipe_config->lane_count, - adjusted_mode->crtc_clock, - pipe_config->port_clock, - &pipe_config->dp_m_n, - constant_n); + if (!pipe_config->dsc_params.compression_enable) + intel_link_compute_m_n(pipe_config->pipe_bpp, + pipe_config->lane_count, + adjusted_mode->crtc_clock, + pipe_config->port_clock, + &pipe_config->dp_m_n, + constant_n); + else + intel_link_compute_m_n(pipe_config->dsc_params.compression_enable, + pipe_config->lane_count, + adjusted_mode->crtc_clock, + pipe_config->port_clock, + &pipe_config->dp_m_n, + constant_n); if (intel_connector->panel.downclock_mode != NULL && dev_priv->drrs.type == SEAMLESS_DRRS_SUPPORT) { From 8228c42fc0ee54dd2c798035baa35739e54bfe91 Mon Sep 17 00:00:00 2001 From: Manasi Navare Date: Wed, 28 Nov 2018 12:26:14 -0800 Subject: [PATCH 36/98] drm/i915/dp: Do not enable PSR2 if DSC is enabled MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit If a eDP panel supports both PSR2 and VDSC, our HW cannot support both at a time. Give priority to PSR2 if a requested resolution can be supported without compression else enable VDSC and keep PSR2 disabled. v4: Fix the unrealted stuff removed during rebase (Ville) v3: * Rebase v2: * Add warning for DSC and PSR2 enabled together (DK) Cc: Rodrigo Vivi Cc: Jani Nikula Cc: Ville Syrjälä Signed-off-by: Manasi Navare Reviewed-by: Rodrigo Vivi Link: https://patchwork.freedesktop.org/patch/msgid/20181128202628.20238-3-manasi.d.navare@intel.com --- drivers/gpu/drm/i915/intel_psr.c | 14 ++++++++++++++ 1 file changed, 14 insertions(+) diff --git a/drivers/gpu/drm/i915/intel_psr.c b/drivers/gpu/drm/i915/intel_psr.c index 572e626eadff..2084784f320d 100644 --- a/drivers/gpu/drm/i915/intel_psr.c +++ b/drivers/gpu/drm/i915/intel_psr.c @@ -75,6 +75,10 @@ static bool intel_psr2_enabled(struct drm_i915_private *dev_priv, if (i915_modparams.enable_psr == -1) return false; + /* Cannot enable DSC and PSR2 simultaneously */ + WARN_ON(crtc_state->dsc_params.compression_enable && + crtc_state->has_psr2); + switch (dev_priv->psr.debug & I915_PSR_DEBUG_MODE_MASK) { case I915_PSR_DEBUG_FORCE_PSR1: return false; @@ -502,6 +506,16 @@ static bool intel_psr2_config_valid(struct intel_dp *intel_dp, if (!dev_priv->psr.sink_psr2_support) return false; + /* + * DSC and PSR2 cannot be enabled simultaneously. If a requested + * resolution requires DSC to be enabled, priority is given to DSC + * over PSR2. + */ + if (crtc_state->dsc_params.compression_enable) { + DRM_DEBUG_KMS("PSR2 cannot be enabled since DSC is enabled\n"); + return false; + } + if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv)) { psr_max_h = 4096; psr_max_v = 2304; From 168243c1801065f7e3a6afb6d5bd49ae5bdf1871 Mon Sep 17 00:00:00 2001 From: Gaurav K Singh Date: Thu, 29 Nov 2018 11:38:27 -0800 Subject: [PATCH 37/98] drm/i915/dsc: Define & Compute VESA DSC params This patches does the following: 1. This patch defines all the DSC parameters as per the VESA DSC specification. These are stored in the encoder and used to compute the PPS parameters to be sent to the Sink. 2. Compute all the DSC parameters which are derived from DSC state of intel_crtc_state. 3. Compute all parameters that are VESA DSC specific This computation happens in the atomic check phase during compute_config() to validate if display stream compression can be enabled for the requested mode. v8 (From Manasi): * DEBUG_KMS instead of DRM_ERROR for user triggerable errors (Ville) v7: (From Manasi) * Dont use signed int for rc_range_params (Manasi) * Mask the range_bpg_offset to use only 6 bits * Add SPDX identifier (Chris Wilson) v6 (From Manasi): * Add a check for line_buf_depth return value (Anusha) * Remove DRM DSC constants to different patch (Manasi) v5 (From Manasi): * Add logic to limit the max line buf depth for DSC 1.1 to 13 as per DSC 1.1 spec * Fix dim checkpatch warnings/checks v4 (From Gaurav): * Rebase on latest drm tip * rename variable name(Manasi) * Populate linebuf_depth variable(Manasi) v3 (From Gaurav): * Rebase my previous patches on top of Manasi's latest patch series * Using >>n rather than /2^n (Manasi) * Change the commit message to explain what the patch is doing(Gaurav) Fixed review comments from Ville: * Don't use macro TWOS_COMPLEMENT * Mention in comment about the source of RC params * Return directly from case statements * Using single asssignment for assigning rc_range_params * Using < Cc: Ville Syrjala Cc: Anusha Srivatsa Cc: Gaurav K Singh Signed-off-by: Gaurav K Singh Signed-off-by: Manasi Navare Co-developed-by: Manasi Navare Reviewed-by: Anusha Srivatsa Link: https://patchwork.freedesktop.org/patch/msgid/20181129193827.7914-1-manasi.d.navare@intel.com --- drivers/gpu/drm/i915/Makefile | 3 +- drivers/gpu/drm/i915/intel_dp.c | 7 + drivers/gpu/drm/i915/intel_drv.h | 4 + drivers/gpu/drm/i915/intel_vdsc.c | 456 ++++++++++++++++++++++++++++++ 4 files changed, 469 insertions(+), 1 deletion(-) create mode 100644 drivers/gpu/drm/i915/intel_vdsc.c diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile index 0ff878c994e2..8370b9de6e4f 100644 --- a/drivers/gpu/drm/i915/Makefile +++ b/drivers/gpu/drm/i915/Makefile @@ -157,7 +157,8 @@ i915-y += dvo_ch7017.o \ intel_sdvo.o \ intel_tv.o \ vlv_dsi.o \ - vlv_dsi_pll.o + vlv_dsi_pll.o \ + intel_vdsc.o # Post-mortem debug and GPU hang state capture i915-$(CONFIG_DRM_I915_CAPTURE_ERROR) += i915_gpu_error.o diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index a2780733768a..3b612eb134e7 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -1953,6 +1953,13 @@ static bool intel_dp_dsc_compute_config(struct intel_dp *intel_dp, return false; } } + if (intel_dp_compute_dsc_params(intel_dp, pipe_config) < 0) { + DRM_DEBUG_KMS("Cannot compute valid DSC parameters for Input Bpp = %d " + "Compressed BPP = %d\n", + pipe_config->pipe_bpp, + pipe_config->dsc_params.compressed_bpp); + return false; + } pipe_config->dsc_params.compression_enable = true; DRM_DEBUG_KMS("DP DSC computed with Input Bpp = %d " "Compressed Bpp = %d Slice Count = %d\n", diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index 5577f0f31b99..44a5b21799f5 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -1857,6 +1857,10 @@ uint16_t intel_dp_dsc_get_output_bpp(int link_clock, uint8_t lane_count, uint8_t intel_dp_dsc_get_slice_count(struct intel_dp *intel_dp, int mode_clock, int mode_hdisplay); +/* intel_vdsc.c */ +int intel_dp_compute_dsc_params(struct intel_dp *intel_dp, + struct intel_crtc_state *pipe_config); + static inline unsigned int intel_dp_unused_lane_mask(int lane_count) { return ~((1 << lane_count) - 1) & 0xf; diff --git a/drivers/gpu/drm/i915/intel_vdsc.c b/drivers/gpu/drm/i915/intel_vdsc.c new file mode 100644 index 000000000000..5e0f9ac5667b --- /dev/null +++ b/drivers/gpu/drm/i915/intel_vdsc.c @@ -0,0 +1,456 @@ +// SPDX-License-Identifier: MIT +/* + * Copyright © 2018 Intel Corporation + * + * Author: Gaurav K Singh + * Manasi Navare + */ + +#include +#include +#include "i915_drv.h" +#include "intel_drv.h" + +enum ROW_INDEX_BPP { + ROW_INDEX_6BPP = 0, + ROW_INDEX_8BPP, + ROW_INDEX_10BPP, + ROW_INDEX_12BPP, + ROW_INDEX_15BPP, + MAX_ROW_INDEX +}; + +enum COLUMN_INDEX_BPC { + COLUMN_INDEX_8BPC = 0, + COLUMN_INDEX_10BPC, + COLUMN_INDEX_12BPC, + COLUMN_INDEX_14BPC, + COLUMN_INDEX_16BPC, + MAX_COLUMN_INDEX +}; + +#define DSC_SUPPORTED_VERSION_MIN 1 + +/* From DSC_v1.11 spec, rc_parameter_Set syntax element typically constant */ +static u16 rc_buf_thresh[] = { + 896, 1792, 2688, 3584, 4480, 5376, 6272, 6720, 7168, 7616, + 7744, 7872, 8000, 8064 +}; + +struct rc_parameters { + u16 initial_xmit_delay; + u8 first_line_bpg_offset; + u16 initial_offset; + u8 flatness_min_qp; + u8 flatness_max_qp; + u8 rc_quant_incr_limit0; + u8 rc_quant_incr_limit1; + struct drm_dsc_rc_range_parameters rc_range_params[DSC_NUM_BUF_RANGES]; +}; + +/* + * Selected Rate Control Related Parameter Recommended Values + * from DSC_v1.11 spec & C Model release: DSC_model_20161212 + */ +static struct rc_parameters rc_params[][MAX_COLUMN_INDEX] = { +{ + /* 6BPP/8BPC */ + { 768, 15, 6144, 3, 13, 11, 11, { + { 0, 4, 0 }, { 1, 6, -2 }, { 3, 8, -2 }, { 4, 8, -4 }, + { 5, 9, -6 }, { 5, 9, -6 }, { 6, 9, -6 }, { 6, 10, -8 }, + { 7, 11, -8 }, { 8, 12, -10 }, { 9, 12, -10 }, { 10, 12, -12 }, + { 10, 12, -12 }, { 11, 12, -12 }, { 13, 14, -12 } + } + }, + /* 6BPP/10BPC */ + { 768, 15, 6144, 7, 17, 15, 15, { + { 0, 8, 0 }, { 3, 10, -2 }, { 7, 12, -2 }, { 8, 12, -4 }, + { 9, 13, -6 }, { 9, 13, -6 }, { 10, 13, -6 }, { 10, 14, -8 }, + { 11, 15, -8 }, { 12, 16, -10 }, { 13, 16, -10 }, + { 14, 16, -12 }, { 14, 16, -12 }, { 15, 16, -12 }, + { 17, 18, -12 } + } + }, + /* 6BPP/12BPC */ + { 768, 15, 6144, 11, 21, 19, 19, { + { 0, 12, 0 }, { 5, 14, -2 }, { 11, 16, -2 }, { 12, 16, -4 }, + { 13, 17, -6 }, { 13, 17, -6 }, { 14, 17, -6 }, { 14, 18, -8 }, + { 15, 19, -8 }, { 16, 20, -10 }, { 17, 20, -10 }, + { 18, 20, -12 }, { 18, 20, -12 }, { 19, 20, -12 }, + { 21, 22, -12 } + } + }, + /* 6BPP/14BPC */ + { 768, 15, 6144, 15, 25, 23, 27, { + { 0, 16, 0 }, { 7, 18, -2 }, { 15, 20, -2 }, { 16, 20, -4 }, + { 17, 21, -6 }, { 17, 21, -6 }, { 18, 21, -6 }, { 18, 22, -8 }, + { 19, 23, -8 }, { 20, 24, -10 }, { 21, 24, -10 }, + { 22, 24, -12 }, { 22, 24, -12 }, { 23, 24, -12 }, + { 25, 26, -12 } + } + }, + /* 6BPP/16BPC */ + { 768, 15, 6144, 19, 29, 27, 27, { + { 0, 20, 0 }, { 9, 22, -2 }, { 19, 24, -2 }, { 20, 24, -4 }, + { 21, 25, -6 }, { 21, 25, -6 }, { 22, 25, -6 }, { 22, 26, -8 }, + { 23, 27, -8 }, { 24, 28, -10 }, { 25, 28, -10 }, + { 26, 28, -12 }, { 26, 28, -12 }, { 27, 28, -12 }, + { 29, 30, -12 } + } + }, +}, +{ + /* 8BPP/8BPC */ + { 512, 12, 6144, 3, 12, 11, 11, { + { 0, 4, 2 }, { 0, 4, 0 }, { 1, 5, 0 }, { 1, 6, -2 }, + { 3, 7, -4 }, { 3, 7, -6 }, { 3, 7, -8 }, { 3, 8, -8 }, + { 3, 9, -8 }, { 3, 10, -10 }, { 5, 11, -10 }, { 5, 12, -12 }, + { 5, 13, -12 }, { 7, 13, -12 }, { 13, 15, -12 } + } + }, + /* 8BPP/10BPC */ + { 512, 12, 6144, 7, 16, 15, 15, { + { 0, 4, 2 }, { 4, 8, 0 }, { 5, 9, 0 }, { 5, 10, -2 }, + { 7, 11, -4 }, { 7, 11, -6 }, { 7, 11, -8 }, { 7, 12, -8 }, + { 7, 13, -8 }, { 7, 14, -10 }, { 9, 15, -10 }, { 9, 16, -12 }, + { 9, 17, -12 }, { 11, 17, -12 }, { 17, 19, -12 } + } + }, + /* 8BPP/12BPC */ + { 512, 12, 6144, 11, 20, 19, 19, { + { 0, 12, 2 }, { 4, 12, 0 }, { 9, 13, 0 }, { 9, 14, -2 }, + { 11, 15, -4 }, { 11, 15, -6 }, { 11, 15, -8 }, { 11, 16, -8 }, + { 11, 17, -8 }, { 11, 18, -10 }, { 13, 19, -10 }, + { 13, 20, -12 }, { 13, 21, -12 }, { 15, 21, -12 }, + { 21, 23, -12 } + } + }, + /* 8BPP/14BPC */ + { 512, 12, 6144, 15, 24, 23, 23, { + { 0, 12, 0 }, { 5, 13, 0 }, { 11, 15, 0 }, { 12, 17, -2 }, + { 15, 19, -4 }, { 15, 19, -6 }, { 15, 19, -8 }, { 15, 20, -8 }, + { 15, 21, -8 }, { 15, 22, -10 }, { 17, 22, -10 }, + { 17, 23, -12 }, { 17, 23, -12 }, { 21, 24, -12 }, + { 24, 25, -12 } + } + }, + /* 8BPP/16BPC */ + { 512, 12, 6144, 19, 28, 27, 27, { + { 0, 12, 2 }, { 6, 14, 0 }, { 13, 17, 0 }, { 15, 20, -2 }, + { 19, 23, -4 }, { 19, 23, -6 }, { 19, 23, -8 }, { 19, 24, -8 }, + { 19, 25, -8 }, { 19, 26, -10 }, { 21, 26, -10 }, + { 21, 27, -12 }, { 21, 27, -12 }, { 25, 28, -12 }, + { 28, 29, -12 } + } + }, +}, +{ + /* 10BPP/8BPC */ + { 410, 15, 5632, 3, 12, 11, 11, { + { 0, 3, 2 }, { 0, 4, 0 }, { 1, 5, 0 }, { 2, 6, -2 }, + { 3, 7, -4 }, { 3, 7, -6 }, { 3, 7, -8 }, { 3, 8, -8 }, + { 3, 9, -8 }, { 3, 9, -10 }, { 5, 10, -10 }, { 5, 10, -10 }, + { 5, 11, -12 }, { 7, 11, -12 }, { 11, 12, -12 } + } + }, + /* 10BPP/10BPC */ + { 410, 15, 5632, 7, 16, 15, 15, { + { 0, 7, 2 }, { 4, 8, 0 }, { 5, 9, 0 }, { 6, 10, -2 }, + { 7, 11, -4 }, { 7, 11, -6 }, { 7, 11, -8 }, { 7, 12, -8 }, + { 7, 13, -8 }, { 7, 13, -10 }, { 9, 14, -10 }, { 9, 14, -10 }, + { 9, 15, -12 }, { 11, 15, -12 }, { 15, 16, -12 } + } + }, + /* 10BPP/12BPC */ + { 410, 15, 5632, 11, 20, 19, 19, { + { 0, 11, 2 }, { 4, 12, 0 }, { 9, 13, 0 }, { 10, 14, -2 }, + { 11, 15, -4 }, { 11, 15, -6 }, { 11, 15, -8 }, { 11, 16, -8 }, + { 11, 17, -8 }, { 11, 17, -10 }, { 13, 18, -10 }, + { 13, 18, -10 }, { 13, 19, -12 }, { 15, 19, -12 }, + { 19, 20, -12 } + } + }, + /* 10BPP/14BPC */ + { 410, 15, 5632, 15, 24, 23, 23, { + { 0, 11, 2 }, { 5, 13, 0 }, { 11, 15, 0 }, { 13, 18, -2 }, + { 15, 19, -4 }, { 15, 19, -6 }, { 15, 19, -8 }, { 15, 20, -8 }, + { 15, 21, -8 }, { 15, 21, -10 }, { 17, 22, -10 }, + { 17, 22, -10 }, { 17, 23, -12 }, { 19, 23, -12 }, + { 23, 24, -12 } + } + }, + /* 10BPP/16BPC */ + { 410, 15, 5632, 19, 28, 27, 27, { + { 0, 11, 2 }, { 6, 14, 0 }, { 13, 17, 0 }, { 16, 20, -2 }, + { 19, 23, -4 }, { 19, 23, -6 }, { 19, 23, -8 }, { 19, 24, -8 }, + { 19, 25, -8 }, { 19, 25, -10 }, { 21, 26, -10 }, + { 21, 26, -10 }, { 21, 27, -12 }, { 23, 27, -12 }, + { 27, 28, -12 } + } + }, +}, +{ + /* 12BPP/8BPC */ + { 341, 15, 2048, 3, 12, 11, 11, { + { 0, 2, 2 }, { 0, 4, 0 }, { 1, 5, 0 }, { 1, 6, -2 }, + { 3, 7, -4 }, { 3, 7, -6 }, { 3, 7, -8 }, { 3, 8, -8 }, + { 3, 9, -8 }, { 3, 10, -10 }, { 5, 11, -10 }, + { 5, 12, -12 }, { 5, 13, -12 }, { 7, 13, -12 }, { 13, 15, -12 } + } + }, + /* 12BPP/10BPC */ + { 341, 15, 2048, 7, 16, 15, 15, { + { 0, 2, 2 }, { 2, 5, 0 }, { 3, 7, 0 }, { 4, 8, -2 }, + { 6, 9, -4 }, { 7, 10, -6 }, { 7, 11, -8 }, { 7, 12, -8 }, + { 7, 13, -8 }, { 7, 14, -10 }, { 9, 15, -10 }, { 9, 16, -12 }, + { 9, 17, -12 }, { 11, 17, -12 }, { 17, 19, -12 } + } + }, + /* 12BPP/12BPC */ + { 341, 15, 2048, 11, 20, 19, 19, { + { 0, 6, 2 }, { 4, 9, 0 }, { 7, 11, 0 }, { 8, 12, -2 }, + { 10, 13, -4 }, { 11, 14, -6 }, { 11, 15, -8 }, { 11, 16, -8 }, + { 11, 17, -8 }, { 11, 18, -10 }, { 13, 19, -10 }, + { 13, 20, -12 }, { 13, 21, -12 }, { 15, 21, -12 }, + { 21, 23, -12 } + } + }, + /* 12BPP/14BPC */ + { 341, 15, 2048, 15, 24, 23, 23, { + { 0, 6, 2 }, { 7, 10, 0 }, { 9, 13, 0 }, { 11, 16, -2 }, + { 14, 17, -4 }, { 15, 18, -6 }, { 15, 19, -8 }, { 15, 20, -8 }, + { 15, 20, -8 }, { 15, 21, -10 }, { 17, 21, -10 }, + { 17, 21, -12 }, { 17, 21, -12 }, { 19, 22, -12 }, + { 22, 23, -12 } + } + }, + /* 12BPP/16BPC */ + { 341, 15, 2048, 19, 28, 27, 27, { + { 0, 6, 2 }, { 6, 11, 0 }, { 11, 15, 0 }, { 14, 18, -2 }, + { 18, 21, -4 }, { 19, 22, -6 }, { 19, 23, -8 }, { 19, 24, -8 }, + { 19, 24, -8 }, { 19, 25, -10 }, { 21, 25, -10 }, + { 21, 25, -12 }, { 21, 25, -12 }, { 23, 26, -12 }, + { 26, 27, -12 } + } + }, +}, +{ + /* 15BPP/8BPC */ + { 273, 15, 2048, 3, 12, 11, 11, { + { 0, 0, 10 }, { 0, 1, 8 }, { 0, 1, 6 }, { 0, 2, 4 }, + { 1, 2, 2 }, { 1, 3, 0 }, { 1, 3, -2 }, { 2, 4, -4 }, + { 2, 5, -6 }, { 3, 5, -8 }, { 4, 6, -10 }, { 4, 7, -10 }, + { 5, 7, -12 }, { 7, 8, -12 }, { 8, 9, -12 } + } + }, + /* 15BPP/10BPC */ + { 273, 15, 2048, 7, 16, 15, 15, { + { 0, 2, 10 }, { 2, 5, 8 }, { 3, 5, 6 }, { 4, 6, 4 }, + { 5, 6, 2 }, { 5, 7, 0 }, { 5, 7, -2 }, { 6, 8, -4 }, + { 6, 9, -6 }, { 7, 9, -8 }, { 8, 10, -10 }, { 8, 11, -10 }, + { 9, 11, -12 }, { 11, 12, -12 }, { 12, 13, -12 } + } + }, + /* 15BPP/12BPC */ + { 273, 15, 2048, 11, 20, 19, 19, { + { 0, 4, 10 }, { 2, 7, 8 }, { 4, 9, 6 }, { 6, 11, 4 }, + { 9, 11, 2 }, { 9, 11, 0 }, { 9, 12, -2 }, { 10, 12, -4 }, + { 11, 13, -6 }, { 11, 13, -8 }, { 12, 14, -10 }, + { 13, 15, -10 }, { 13, 15, -12 }, { 15, 16, -12 }, + { 16, 17, -12 } + } + }, + /* 15BPP/14BPC */ + { 273, 15, 2048, 15, 24, 23, 23, { + { 0, 4, 10 }, { 3, 8, 8 }, { 6, 11, 6 }, { 9, 14, 4 }, + { 13, 15, 2 }, { 13, 15, 0 }, { 13, 16, -2 }, { 14, 16, -4 }, + { 15, 17, -6 }, { 15, 17, -8 }, { 16, 18, -10 }, + { 17, 19, -10 }, { 17, 19, -12 }, { 19, 20, -12 }, + { 20, 21, -12 } + } + }, + /* 15BPP/16BPC */ + { 273, 15, 2048, 19, 28, 27, 27, { + { 0, 4, 10 }, { 4, 9, 8 }, { 8, 13, 6 }, { 12, 17, 4 }, + { 17, 19, 2 }, { 17, 20, 0 }, { 17, 20, -2 }, { 18, 20, -4 }, + { 19, 21, -6 }, { 19, 21, -8 }, { 20, 22, -10 }, + { 21, 23, -10 }, { 21, 23, -12 }, { 23, 24, -12 }, + { 24, 25, -12 } + } + } +} + +}; + +static int get_row_index_for_rc_params(u16 compressed_bpp) +{ + switch (compressed_bpp) { + case 6: + return ROW_INDEX_6BPP; + case 8: + return ROW_INDEX_8BPP; + case 10: + return ROW_INDEX_10BPP; + case 12: + return ROW_INDEX_12BPP; + case 15: + return ROW_INDEX_15BPP; + default: + return -EINVAL; + } +} + +static int get_column_index_for_rc_params(u8 bits_per_component) +{ + switch (bits_per_component) { + case 8: + return COLUMN_INDEX_8BPC; + case 10: + return COLUMN_INDEX_10BPC; + case 12: + return COLUMN_INDEX_12BPC; + case 14: + return COLUMN_INDEX_14BPC; + case 16: + return COLUMN_INDEX_16BPC; + default: + return -EINVAL; + } +} + +int intel_dp_compute_dsc_params(struct intel_dp *intel_dp, + struct intel_crtc_state *pipe_config) +{ + struct drm_dsc_config *vdsc_cfg = &pipe_config->dp_dsc_cfg; + u16 compressed_bpp = pipe_config->dsc_params.compressed_bpp; + u8 i = 0; + int row_index = 0; + int column_index = 0; + u8 line_buf_depth = 0; + + vdsc_cfg->pic_width = pipe_config->base.adjusted_mode.crtc_hdisplay; + vdsc_cfg->pic_height = pipe_config->base.adjusted_mode.crtc_vdisplay; + vdsc_cfg->slice_width = DIV_ROUND_UP(vdsc_cfg->pic_width, + pipe_config->dsc_params.slice_count); + /* + * Slice Height of 8 works for all currently available panels. So start + * with that if pic_height is an integral multiple of 8. + * Eventually add logic to try multiple slice heights. + */ + if (vdsc_cfg->pic_height % 8 == 0) + vdsc_cfg->slice_height = 8; + else if (vdsc_cfg->pic_height % 4 == 0) + vdsc_cfg->slice_height = 4; + else + vdsc_cfg->slice_height = 2; + + /* Values filled from DSC Sink DPCD */ + vdsc_cfg->dsc_version_major = + (intel_dp->dsc_dpcd[DP_DSC_REV - DP_DSC_SUPPORT] & + DP_DSC_MAJOR_MASK) >> DP_DSC_MAJOR_SHIFT; + vdsc_cfg->dsc_version_minor = + min(DSC_SUPPORTED_VERSION_MIN, + (intel_dp->dsc_dpcd[DP_DSC_REV - DP_DSC_SUPPORT] & + DP_DSC_MINOR_MASK) >> DP_DSC_MINOR_SHIFT); + + vdsc_cfg->convert_rgb = intel_dp->dsc_dpcd[DP_DSC_DEC_COLOR_FORMAT_CAP - DP_DSC_SUPPORT] & + DP_DSC_RGB; + + line_buf_depth = drm_dp_dsc_sink_line_buf_depth(intel_dp->dsc_dpcd); + if (!line_buf_depth) { + DRM_DEBUG_KMS("DSC Sink Line Buffer Depth invalid\n"); + return -EINVAL; + } + if (vdsc_cfg->dsc_version_minor == 2) + vdsc_cfg->line_buf_depth = (line_buf_depth == DSC_1_2_MAX_LINEBUF_DEPTH_BITS) ? + DSC_1_2_MAX_LINEBUF_DEPTH_VAL : line_buf_depth; + else + vdsc_cfg->line_buf_depth = (line_buf_depth > DSC_1_1_MAX_LINEBUF_DEPTH_BITS) ? + DSC_1_1_MAX_LINEBUF_DEPTH_BITS : line_buf_depth; + + /* Gen 11 does not support YCbCr */ + vdsc_cfg->enable422 = false; + /* Gen 11 does not support VBR */ + vdsc_cfg->vbr_enable = false; + vdsc_cfg->block_pred_enable = + intel_dp->dsc_dpcd[DP_DSC_BLK_PREDICTION_SUPPORT - DP_DSC_SUPPORT] & + DP_DSC_BLK_PREDICTION_IS_SUPPORTED; + + /* Gen 11 only supports integral values of bpp */ + vdsc_cfg->bits_per_pixel = compressed_bpp << 4; + vdsc_cfg->bits_per_component = pipe_config->pipe_bpp / 3; + + for (i = 0; i < DSC_NUM_BUF_RANGES - 1; i++) { + /* + * six 0s are appended to the lsb of each threshold value + * internally in h/w. + * Only 8 bits are allowed for programming RcBufThreshold + */ + vdsc_cfg->rc_buf_thresh[i] = rc_buf_thresh[i] >> 6; + } + + /* + * For 6bpp, RC Buffer threshold 12 and 13 need a different value + * as per C Model + */ + if (compressed_bpp == 6) { + vdsc_cfg->rc_buf_thresh[12] = 0x7C; + vdsc_cfg->rc_buf_thresh[13] = 0x7D; + } + + row_index = get_row_index_for_rc_params(compressed_bpp); + column_index = + get_column_index_for_rc_params(vdsc_cfg->bits_per_component); + + if (row_index < 0 || column_index < 0) + return -EINVAL; + + vdsc_cfg->first_line_bpg_offset = + rc_params[row_index][column_index].first_line_bpg_offset; + vdsc_cfg->initial_xmit_delay = + rc_params[row_index][column_index].initial_xmit_delay; + vdsc_cfg->initial_offset = + rc_params[row_index][column_index].initial_offset; + vdsc_cfg->flatness_min_qp = + rc_params[row_index][column_index].flatness_min_qp; + vdsc_cfg->flatness_max_qp = + rc_params[row_index][column_index].flatness_max_qp; + vdsc_cfg->rc_quant_incr_limit0 = + rc_params[row_index][column_index].rc_quant_incr_limit0; + vdsc_cfg->rc_quant_incr_limit1 = + rc_params[row_index][column_index].rc_quant_incr_limit1; + + for (i = 0; i < DSC_NUM_BUF_RANGES; i++) { + vdsc_cfg->rc_range_params[i].range_min_qp = + rc_params[row_index][column_index].rc_range_params[i].range_min_qp; + vdsc_cfg->rc_range_params[i].range_max_qp = + rc_params[row_index][column_index].rc_range_params[i].range_max_qp; + /* + * Range BPG Offset uses 2's complement and is only a 6 bits. So + * mask it to get only 6 bits. + */ + vdsc_cfg->rc_range_params[i].range_bpg_offset = + rc_params[row_index][column_index].rc_range_params[i].range_bpg_offset & + DSC_RANGE_BPG_OFFSET_MASK; + } + + /* + * BitsPerComponent value determines mux_word_size: + * When BitsPerComponent is 12bpc, muxWordSize will be equal to 64 bits + * When BitsPerComponent is 8 or 10bpc, muxWordSize will be equal to + * 48 bits + */ + if (vdsc_cfg->bits_per_component == 8 || + vdsc_cfg->bits_per_component == 10) + vdsc_cfg->mux_word_size = DSC_MUX_WORD_SIZE_8_10_BPC; + else if (vdsc_cfg->bits_per_component == 12) + vdsc_cfg->mux_word_size = DSC_MUX_WORD_SIZE_12_BPC; + + /* RC_MODEL_SIZE is a constant across all configurations */ + vdsc_cfg->rc_model_size = DSC_RC_MODEL_SIZE_CONST; + /* InitialScaleValue is a 6 bit value with 3 fractional bits (U3.3) */ + vdsc_cfg->initial_scale_value = (vdsc_cfg->rc_model_size << 3) / + (vdsc_cfg->rc_model_size - vdsc_cfg->initial_offset); + + return 0; +} From 27998631458a377bd1845f39b65492605299ad08 Mon Sep 17 00:00:00 2001 From: Gaurav K Singh Date: Wed, 28 Nov 2018 12:26:16 -0800 Subject: [PATCH 38/98] drm/i915/dsc: Compute Rate Control parameters for DSC This computation of RC params happens in the atomic commit phase during compute_config() to validate if display stream compression can be enabled for the requested mode. v7 (From Manasi): * Use DRM_DEBUG instead of DRM_ERROR (Ville) * Use Error numberinstead of -1 (Ville) v6 (From Manasi): * Use 9 instead of 0x9 for consistency (Anusha) v5 (From Manasi): * Fix dim checkpatch warnings/checks v4(From Gaurav): * No change.Rebase on drm-tip v3 (From Gaurav): * Rebase on top of Manasi's latest series * Return -ve value in case of failure scenarios (Manasi) Fix review comments from Ville: * Remove unnecessary comments * Remove unnecessary paranthesis * Add comments for few RC params calculations v2 (From Manasi): * Rebase Gaurav's patch from intel-gfx to gfx-internal * Use struct drm_dsc_cfg instead of struct intel_dp as a parameter Cc: Manasi Navare Cc: Jani Nikula Cc: Ville Syrjala Signed-off-by: Gaurav K Singh Signed-off-by: Manasi Navare Reviewed-by: Anusha Srivatsa Link: https://patchwork.freedesktop.org/patch/msgid/20181128202628.20238-5-manasi.d.navare@intel.com --- drivers/gpu/drm/i915/intel_vdsc.c | 125 +++++++++++++++++++++++++++++- 1 file changed, 124 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/intel_vdsc.c b/drivers/gpu/drm/i915/intel_vdsc.c index 5e0f9ac5667b..fde1f184a30b 100644 --- a/drivers/gpu/drm/i915/intel_vdsc.c +++ b/drivers/gpu/drm/i915/intel_vdsc.c @@ -318,6 +318,129 @@ static int get_column_index_for_rc_params(u8 bits_per_component) } } +static int intel_compute_rc_parameters(struct drm_dsc_config *vdsc_cfg) +{ + unsigned long groups_per_line = 0; + unsigned long groups_total = 0; + unsigned long num_extra_mux_bits = 0; + unsigned long slice_bits = 0; + unsigned long hrd_delay = 0; + unsigned long final_scale = 0; + unsigned long rbs_min = 0; + + /* Number of groups used to code each line of a slice */ + groups_per_line = DIV_ROUND_UP(vdsc_cfg->slice_width, + DSC_RC_PIXELS_PER_GROUP); + + /* chunksize in Bytes */ + vdsc_cfg->slice_chunk_size = DIV_ROUND_UP(vdsc_cfg->slice_width * + vdsc_cfg->bits_per_pixel, + (8 * 16)); + + if (vdsc_cfg->convert_rgb) + num_extra_mux_bits = 3 * (vdsc_cfg->mux_word_size + + (4 * vdsc_cfg->bits_per_component + 4) + - 2); + else + num_extra_mux_bits = 3 * vdsc_cfg->mux_word_size + + (4 * vdsc_cfg->bits_per_component + 4) + + 2 * (4 * vdsc_cfg->bits_per_component) - 2; + /* Number of bits in one Slice */ + slice_bits = 8 * vdsc_cfg->slice_chunk_size * vdsc_cfg->slice_height; + + while ((num_extra_mux_bits > 0) && + ((slice_bits - num_extra_mux_bits) % vdsc_cfg->mux_word_size)) + num_extra_mux_bits--; + + if (groups_per_line < vdsc_cfg->initial_scale_value - 8) + vdsc_cfg->initial_scale_value = groups_per_line + 8; + + /* scale_decrement_interval calculation according to DSC spec 1.11 */ + if (vdsc_cfg->initial_scale_value > 8) + vdsc_cfg->scale_decrement_interval = groups_per_line / + (vdsc_cfg->initial_scale_value - 8); + else + vdsc_cfg->scale_decrement_interval = DSC_SCALE_DECREMENT_INTERVAL_MAX; + + vdsc_cfg->final_offset = vdsc_cfg->rc_model_size - + (vdsc_cfg->initial_xmit_delay * + vdsc_cfg->bits_per_pixel + 8) / 16 + num_extra_mux_bits; + + if (vdsc_cfg->final_offset >= vdsc_cfg->rc_model_size) { + DRM_DEBUG_KMS("FinalOfs < RcModelSze for this InitialXmitDelay\n"); + return -ERANGE; + } + + final_scale = (vdsc_cfg->rc_model_size * 8) / + (vdsc_cfg->rc_model_size - vdsc_cfg->final_offset); + if (vdsc_cfg->slice_height > 1) + /* + * NflBpgOffset is 16 bit value with 11 fractional bits + * hence we multiply by 2^11 for preserving the + * fractional part + */ + vdsc_cfg->nfl_bpg_offset = DIV_ROUND_UP((vdsc_cfg->first_line_bpg_offset << 11), + (vdsc_cfg->slice_height - 1)); + else + vdsc_cfg->nfl_bpg_offset = 0; + + /* 2^16 - 1 */ + if (vdsc_cfg->nfl_bpg_offset > 65535) { + DRM_DEBUG_KMS("NflBpgOffset is too large for this slice height\n"); + return -ERANGE; + } + + /* Number of groups used to code the entire slice */ + groups_total = groups_per_line * vdsc_cfg->slice_height; + + /* slice_bpg_offset is 16 bit value with 11 fractional bits */ + vdsc_cfg->slice_bpg_offset = DIV_ROUND_UP(((vdsc_cfg->rc_model_size - + vdsc_cfg->initial_offset + + num_extra_mux_bits) << 11), + groups_total); + + if (final_scale > 9) { + /* + * ScaleIncrementInterval = + * finaloffset/((NflBpgOffset + SliceBpgOffset)*8(finalscale - 1.125)) + * as (NflBpgOffset + SliceBpgOffset) has 11 bit fractional value, + * we need divide by 2^11 from pstDscCfg values + */ + vdsc_cfg->scale_increment_interval = + (vdsc_cfg->final_offset * (1 << 11)) / + ((vdsc_cfg->nfl_bpg_offset + + vdsc_cfg->slice_bpg_offset) * + (final_scale - 9)); + } else { + /* + * If finalScaleValue is less than or equal to 9, a value of 0 should + * be used to disable the scale increment at the end of the slice + */ + vdsc_cfg->scale_increment_interval = 0; + } + + if (vdsc_cfg->scale_increment_interval > 65535) { + DRM_DEBUG_KMS("ScaleIncrementInterval is large for slice height\n"); + return -ERANGE; + } + + /* + * DSC spec mentions that bits_per_pixel specifies the target + * bits/pixel (bpp) rate that is used by the encoder, + * in steps of 1/16 of a bit per pixel + */ + rbs_min = vdsc_cfg->rc_model_size - vdsc_cfg->initial_offset + + DIV_ROUND_UP(vdsc_cfg->initial_xmit_delay * + vdsc_cfg->bits_per_pixel, 16) + + groups_per_line * vdsc_cfg->first_line_bpg_offset; + + hrd_delay = DIV_ROUND_UP((rbs_min * 16), vdsc_cfg->bits_per_pixel); + vdsc_cfg->rc_bits = (hrd_delay * vdsc_cfg->bits_per_pixel) / 16; + vdsc_cfg->initial_dec_delay = hrd_delay - vdsc_cfg->initial_xmit_delay; + + return 0; +} + int intel_dp_compute_dsc_params(struct intel_dp *intel_dp, struct intel_crtc_state *pipe_config) { @@ -452,5 +575,5 @@ int intel_dp_compute_dsc_params(struct intel_dp *intel_dp, vdsc_cfg->initial_scale_value = (vdsc_cfg->rc_model_size << 3) / (vdsc_cfg->rc_model_size - vdsc_cfg->initial_offset); - return 0; + return intel_compute_rc_parameters(vdsc_cfg); } From 2279298dbf367c2f581cc61840a2e1700e9ceca0 Mon Sep 17 00:00:00 2001 From: Gaurav K Singh Date: Wed, 28 Nov 2018 12:26:17 -0800 Subject: [PATCH 39/98] drm/i915/dp: Enable/Disable DSC in DP Sink This patch enables decompression support in sink device before link training and disables the same during the DDI disabling. v3 (From manasi): * Pass bool state to enable/disable (Ville) v2:(From Manasi) * Change the enable/disable function to take crtc_state instead of intel_dp as an argument (Manasi) * Use the compression_enable flag as part of crtc_state (Manasi) Cc: Jani Nikula Cc: Ville Syrjala Cc: Anusha Srivatsa Cc: Gaurav K Singh Signed-off-by: Gaurav K Singh Signed-off-by: Manasi Navare Reviewed-by: Anusha Srivatsa Link: https://patchwork.freedesktop.org/patch/msgid/20181128202628.20238-6-manasi.d.navare@intel.com --- drivers/gpu/drm/i915/intel_ddi.c | 5 +++++ drivers/gpu/drm/i915/intel_dp.c | 16 ++++++++++++++++ drivers/gpu/drm/i915/intel_drv.h | 3 +++ 3 files changed, 24 insertions(+) diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c index ad11540ac436..fa5ad62cd0db 100644 --- a/drivers/gpu/drm/i915/intel_ddi.c +++ b/drivers/gpu/drm/i915/intel_ddi.c @@ -3134,6 +3134,8 @@ static void intel_ddi_pre_enable_dp(struct intel_encoder *encoder, intel_ddi_init_dp_buf_reg(encoder); if (!is_mst) intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_ON); + intel_dp_sink_set_decompression_state(intel_dp, crtc_state, + true); intel_dp_start_link_train(intel_dp); if (port != PORT_A || INTEL_GEN(dev_priv) >= 9) intel_dp_stop_link_train(intel_dp); @@ -3491,6 +3493,9 @@ static void intel_disable_ddi_dp(struct intel_encoder *encoder, intel_edp_drrs_disable(intel_dp, old_crtc_state); intel_psr_disable(intel_dp, old_crtc_state); intel_edp_backlight_off(old_conn_state); + /* Disable the decompression in DP Sink */ + intel_dp_sink_set_decompression_state(intel_dp, old_crtc_state, + false); } static void intel_disable_ddi_hdmi(struct intel_encoder *encoder, diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 3b612eb134e7..4e65dd1a6c07 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -2857,6 +2857,22 @@ static bool downstream_hpd_needs_d0(struct intel_dp *intel_dp) intel_dp->downstream_ports[0] & DP_DS_PORT_HPD; } +void intel_dp_sink_set_decompression_state(struct intel_dp *intel_dp, + const struct intel_crtc_state *crtc_state, + bool enable) +{ + int ret; + + if (!crtc_state->dsc_params.compression_enable) + return; + + ret = drm_dp_dpcd_writeb(&intel_dp->aux, DP_DSC_ENABLE, + enable ? DP_DECOMPRESSION_EN : 0); + if (ret < 0) + DRM_DEBUG_KMS("Failed to %s sink decompression state\n", + enable ? "enable" : "disable"); +} + /* If the sink supports it, try to set the power state appropriately */ void intel_dp_sink_dpms(struct intel_dp *intel_dp, int mode) { diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index 44a5b21799f5..36e756e15124 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -1802,6 +1802,9 @@ void intel_dp_stop_link_train(struct intel_dp *intel_dp); int intel_dp_retrain_link(struct intel_encoder *encoder, struct drm_modeset_acquire_ctx *ctx); void intel_dp_sink_dpms(struct intel_dp *intel_dp, int mode); +void intel_dp_sink_set_decompression_state(struct intel_dp *intel_dp, + const struct intel_crtc_state *crtc_state, + bool enable); void intel_dp_encoder_reset(struct drm_encoder *encoder); void intel_dp_encoder_suspend(struct intel_encoder *intel_encoder); void intel_dp_encoder_destroy(struct drm_encoder *encoder); From 91ba2c8be4b7eef3c9c424ddd862cdb302f252f3 Mon Sep 17 00:00:00 2001 From: Manasi Navare Date: Wed, 28 Nov 2018 12:26:18 -0800 Subject: [PATCH 40/98] drm/i915/dsc: Add a power domain for VDSC on eDP/MIPI DSI On Icelake, a separate power well PG2 is created for VDSC engine used for eDP/MIPI DSI. This patch adds a new display power domain for Power well 2. v3: * Call it POWER_DOMAIN_TRANSCODER_EDP_VDSC (Ville) * Move it around TRANSCODER power domain defs (Ville) v2: * Fix the power well mismatch CI error (Ville) * Rename as VDSC_PIPE_A (Imre) * Fix a whitespace (Anusha) * Fix Comments (Imre) Cc: Ville Syrjala Cc: Rodrigo Vivi Cc: Imre Deak Signed-off-by: Manasi Navare Reviewed-by: Ville Syrjala Link: https://patchwork.freedesktop.org/patch/msgid/20181128202628.20238-7-manasi.d.navare@intel.com --- drivers/gpu/drm/i915/intel_display.h | 1 + drivers/gpu/drm/i915/intel_runtime_pm.c | 4 +++- 2 files changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/intel_display.h b/drivers/gpu/drm/i915/intel_display.h index 3ec704c93d6e..4262452963b3 100644 --- a/drivers/gpu/drm/i915/intel_display.h +++ b/drivers/gpu/drm/i915/intel_display.h @@ -242,6 +242,7 @@ enum intel_display_power_domain { POWER_DOMAIN_TRANSCODER_B, POWER_DOMAIN_TRANSCODER_C, POWER_DOMAIN_TRANSCODER_EDP, + POWER_DOMAIN_TRANSCODER_EDP_VDSC, POWER_DOMAIN_TRANSCODER_DSI_A, POWER_DOMAIN_TRANSCODER_DSI_C, POWER_DOMAIN_PORT_DDI_A_LANES, diff --git a/drivers/gpu/drm/i915/intel_runtime_pm.c b/drivers/gpu/drm/i915/intel_runtime_pm.c index 1c2de9b69a19..4350a5270423 100644 --- a/drivers/gpu/drm/i915/intel_runtime_pm.c +++ b/drivers/gpu/drm/i915/intel_runtime_pm.c @@ -76,6 +76,8 @@ intel_display_power_domain_str(enum intel_display_power_domain domain) return "TRANSCODER_C"; case POWER_DOMAIN_TRANSCODER_EDP: return "TRANSCODER_EDP"; + case POWER_DOMAIN_TRANSCODER_EDP_VDSC: + return "TRANSCODER_EDP_VDSC"; case POWER_DOMAIN_TRANSCODER_DSI_A: return "TRANSCODER_DSI_A"; case POWER_DOMAIN_TRANSCODER_DSI_C: @@ -2028,9 +2030,9 @@ void intel_display_power_put(struct drm_i915_private *dev_priv, */ #define ICL_PW_2_POWER_DOMAINS ( \ ICL_PW_3_POWER_DOMAINS | \ + BIT_ULL(POWER_DOMAIN_TRANSCODER_EDP_VDSC) | \ BIT_ULL(POWER_DOMAIN_INIT)) /* - * - eDP/DSI VDSC * - KVMR (HW control) */ #define ICL_DISPLAY_DC_OFF_POWER_DOMAINS ( \ From 7182414e25304d0f8c7071756b6e986227c70872 Mon Sep 17 00:00:00 2001 From: Manasi Navare Date: Wed, 28 Nov 2018 12:26:19 -0800 Subject: [PATCH 41/98] drm/i915/dp: Configure i915 Picture parameter Set registers during DSC enabling After encoder->pre_enable() hook, after link training sequence is completed, PPS registers for DSC encoder are configured using the DSC state parameters in intel_crtc_state as part of DSC enabling routine in the source. DSC enabling routine is called after encoder->pre_enable() before enbaling the pipe and after compression is enabled on the sink. v7: * Remove unnecessary comments, leftovers (Ville) * No need for explicit val &= ~ (Ville) v6: intel_dsc_enable to be part of pre_enable hook (Ville) v5: * make crtc_state const (Ville) v4: * Use cpu_transcoder instead of encoder->type for using EDP transcoder DSC registers(Ville) * Keep all PSS regs together (Anusha) v3: * Configure Pic_width/2 for each VDSC engine when two VDSC engines per pipe are used (Manasi) * Add DSC slice_row_per_frame in PPS16 (Manasi) v2: * Enable PG2 power well for VDSC on eDP Cc: Jani Nikula Cc: Ville Syrjala Cc: Anusha Srivatsa Signed-off-by: Manasi Navare [manasi: fixup the line longer than 100 chars while applying] Reviewed-by: Anusha Srivatsa Link: https://patchwork.freedesktop.org/patch/msgid/20181128202628.20238-8-manasi.d.navare@intel.com --- drivers/gpu/drm/i915/i915_drv.h | 2 + drivers/gpu/drm/i915/intel_ddi.c | 2 + drivers/gpu/drm/i915/intel_vdsc.c | 409 ++++++++++++++++++++++++++++++ 3 files changed, 413 insertions(+) diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 42704e731364..06d1eaeeba7a 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -3340,6 +3340,8 @@ extern void intel_rps_mark_interactive(struct drm_i915_private *i915, bool interactive); extern bool intel_set_memory_cxsr(struct drm_i915_private *dev_priv, bool enable); +void intel_dsc_enable(struct intel_encoder *encoder, + const struct intel_crtc_state *crtc_state); int i915_reg_read_ioctl(struct drm_device *dev, void *data, struct drm_file *file); diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c index fa5ad62cd0db..339be10986d7 100644 --- a/drivers/gpu/drm/i915/intel_ddi.c +++ b/drivers/gpu/drm/i915/intel_ddi.c @@ -3144,6 +3144,8 @@ static void intel_ddi_pre_enable_dp(struct intel_encoder *encoder, if (!is_mst) intel_ddi_enable_pipe_clock(crtc_state); + + intel_dsc_enable(encoder, crtc_state); } static void intel_ddi_pre_enable_hdmi(struct intel_encoder *encoder, diff --git a/drivers/gpu/drm/i915/intel_vdsc.c b/drivers/gpu/drm/i915/intel_vdsc.c index fde1f184a30b..9676945a85bb 100644 --- a/drivers/gpu/drm/i915/intel_vdsc.c +++ b/drivers/gpu/drm/i915/intel_vdsc.c @@ -577,3 +577,412 @@ int intel_dp_compute_dsc_params(struct intel_dp *intel_dp, return intel_compute_rc_parameters(vdsc_cfg); } + +static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder, + const struct intel_crtc_state *crtc_state) +{ + struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc); + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); + const struct drm_dsc_config *vdsc_cfg = &crtc_state->dp_dsc_cfg; + enum pipe pipe = crtc->pipe; + enum transcoder cpu_transcoder = crtc_state->cpu_transcoder; + u32 pps_val = 0; + u32 rc_buf_thresh_dword[4]; + u32 rc_range_params_dword[8]; + u8 num_vdsc_instances = (crtc_state->dsc_params.dsc_split) ? 2 : 1; + int i = 0; + + /* Populate PICTURE_PARAMETER_SET_0 registers */ + pps_val = DSC_VER_MAJ | vdsc_cfg->dsc_version_minor << + DSC_VER_MIN_SHIFT | + vdsc_cfg->bits_per_component << DSC_BPC_SHIFT | + vdsc_cfg->line_buf_depth << DSC_LINE_BUF_DEPTH_SHIFT; + if (vdsc_cfg->block_pred_enable) + pps_val |= DSC_BLOCK_PREDICTION; + if (vdsc_cfg->convert_rgb) + pps_val |= DSC_COLOR_SPACE_CONVERSION; + if (vdsc_cfg->enable422) + pps_val |= DSC_422_ENABLE; + if (vdsc_cfg->vbr_enable) + pps_val |= DSC_VBR_ENABLE; + DRM_INFO("PPS0 = 0x%08x\n", pps_val); + if (cpu_transcoder == TRANSCODER_EDP) { + I915_WRITE(DSCA_PICTURE_PARAMETER_SET_0, pps_val); + /* + * If 2 VDSC instances are needed, configure PPS for second + * VDSC + */ + if (crtc_state->dsc_params.dsc_split) + I915_WRITE(DSCC_PICTURE_PARAMETER_SET_0, pps_val); + } else { + I915_WRITE(ICL_DSC0_PICTURE_PARAMETER_SET_0(pipe), pps_val); + if (crtc_state->dsc_params.dsc_split) + I915_WRITE(ICL_DSC1_PICTURE_PARAMETER_SET_0(pipe), + pps_val); + } + + /* Populate PICTURE_PARAMETER_SET_1 registers */ + pps_val = 0; + pps_val |= DSC_BPP(vdsc_cfg->bits_per_pixel); + DRM_INFO("PPS1 = 0x%08x\n", pps_val); + if (cpu_transcoder == TRANSCODER_EDP) { + I915_WRITE(DSCA_PICTURE_PARAMETER_SET_1, pps_val); + /* + * If 2 VDSC instances are needed, configure PPS for second + * VDSC + */ + if (crtc_state->dsc_params.dsc_split) + I915_WRITE(DSCC_PICTURE_PARAMETER_SET_1, pps_val); + } else { + I915_WRITE(ICL_DSC0_PICTURE_PARAMETER_SET_1(pipe), pps_val); + if (crtc_state->dsc_params.dsc_split) + I915_WRITE(ICL_DSC1_PICTURE_PARAMETER_SET_1(pipe), + pps_val); + } + + /* Populate PICTURE_PARAMETER_SET_2 registers */ + pps_val = 0; + pps_val |= DSC_PIC_HEIGHT(vdsc_cfg->pic_height) | + DSC_PIC_WIDTH(vdsc_cfg->pic_width / num_vdsc_instances); + DRM_INFO("PPS2 = 0x%08x\n", pps_val); + if (encoder->type == INTEL_OUTPUT_EDP) { + I915_WRITE(DSCA_PICTURE_PARAMETER_SET_2, pps_val); + /* + * If 2 VDSC instances are needed, configure PPS for second + * VDSC + */ + if (crtc_state->dsc_params.dsc_split) + I915_WRITE(DSCC_PICTURE_PARAMETER_SET_2, pps_val); + } else { + I915_WRITE(ICL_DSC0_PICTURE_PARAMETER_SET_2(pipe), pps_val); + if (crtc_state->dsc_params.dsc_split) + I915_WRITE(ICL_DSC1_PICTURE_PARAMETER_SET_2(pipe), + pps_val); + } + + /* Populate PICTURE_PARAMETER_SET_3 registers */ + pps_val = 0; + pps_val |= DSC_SLICE_HEIGHT(vdsc_cfg->slice_height) | + DSC_SLICE_WIDTH(vdsc_cfg->slice_width); + DRM_INFO("PPS3 = 0x%08x\n", pps_val); + if (cpu_transcoder == TRANSCODER_EDP) { + I915_WRITE(DSCA_PICTURE_PARAMETER_SET_3, pps_val); + /* + * If 2 VDSC instances are needed, configure PPS for second + * VDSC + */ + if (crtc_state->dsc_params.dsc_split) + I915_WRITE(DSCC_PICTURE_PARAMETER_SET_3, pps_val); + } else { + I915_WRITE(ICL_DSC0_PICTURE_PARAMETER_SET_3(pipe), pps_val); + if (crtc_state->dsc_params.dsc_split) + I915_WRITE(ICL_DSC1_PICTURE_PARAMETER_SET_3(pipe), + pps_val); + } + + /* Populate PICTURE_PARAMETER_SET_4 registers */ + pps_val = 0; + pps_val |= DSC_INITIAL_XMIT_DELAY(vdsc_cfg->initial_xmit_delay) | + DSC_INITIAL_DEC_DELAY(vdsc_cfg->initial_dec_delay); + DRM_INFO("PPS4 = 0x%08x\n", pps_val); + if (cpu_transcoder == TRANSCODER_EDP) { + I915_WRITE(DSCA_PICTURE_PARAMETER_SET_4, pps_val); + /* + * If 2 VDSC instances are needed, configure PPS for second + * VDSC + */ + if (crtc_state->dsc_params.dsc_split) + I915_WRITE(DSCC_PICTURE_PARAMETER_SET_4, pps_val); + } else { + I915_WRITE(ICL_DSC0_PICTURE_PARAMETER_SET_4(pipe), pps_val); + if (crtc_state->dsc_params.dsc_split) + I915_WRITE(ICL_DSC1_PICTURE_PARAMETER_SET_4(pipe), + pps_val); + } + + /* Populate PICTURE_PARAMETER_SET_5 registers */ + pps_val = 0; + pps_val |= DSC_SCALE_INC_INT(vdsc_cfg->scale_increment_interval) | + DSC_SCALE_DEC_INT(vdsc_cfg->scale_decrement_interval); + DRM_INFO("PPS5 = 0x%08x\n", pps_val); + if (cpu_transcoder == TRANSCODER_EDP) { + I915_WRITE(DSCA_PICTURE_PARAMETER_SET_5, pps_val); + /* + * If 2 VDSC instances are needed, configure PPS for second + * VDSC + */ + if (crtc_state->dsc_params.dsc_split) + I915_WRITE(DSCC_PICTURE_PARAMETER_SET_5, pps_val); + } else { + I915_WRITE(ICL_DSC0_PICTURE_PARAMETER_SET_5(pipe), pps_val); + if (crtc_state->dsc_params.dsc_split) + I915_WRITE(ICL_DSC1_PICTURE_PARAMETER_SET_5(pipe), + pps_val); + } + + /* Populate PICTURE_PARAMETER_SET_6 registers */ + pps_val = 0; + pps_val |= DSC_INITIAL_SCALE_VALUE(vdsc_cfg->initial_scale_value) | + DSC_FIRST_LINE_BPG_OFFSET(vdsc_cfg->first_line_bpg_offset) | + DSC_FLATNESS_MIN_QP(vdsc_cfg->flatness_min_qp) | + DSC_FLATNESS_MAX_QP(vdsc_cfg->flatness_max_qp); + DRM_INFO("PPS6 = 0x%08x\n", pps_val); + if (cpu_transcoder == TRANSCODER_EDP) { + I915_WRITE(DSCA_PICTURE_PARAMETER_SET_6, pps_val); + /* + * If 2 VDSC instances are needed, configure PPS for second + * VDSC + */ + if (crtc_state->dsc_params.dsc_split) + I915_WRITE(DSCC_PICTURE_PARAMETER_SET_6, pps_val); + } else { + I915_WRITE(ICL_DSC0_PICTURE_PARAMETER_SET_6(pipe), pps_val); + if (crtc_state->dsc_params.dsc_split) + I915_WRITE(ICL_DSC1_PICTURE_PARAMETER_SET_6(pipe), + pps_val); + } + + /* Populate PICTURE_PARAMETER_SET_7 registers */ + pps_val = 0; + pps_val |= DSC_SLICE_BPG_OFFSET(vdsc_cfg->slice_bpg_offset) | + DSC_NFL_BPG_OFFSET(vdsc_cfg->nfl_bpg_offset); + DRM_INFO("PPS7 = 0x%08x\n", pps_val); + if (cpu_transcoder == TRANSCODER_EDP) { + I915_WRITE(DSCA_PICTURE_PARAMETER_SET_7, pps_val); + /* + * If 2 VDSC instances are needed, configure PPS for second + * VDSC + */ + if (crtc_state->dsc_params.dsc_split) + I915_WRITE(DSCC_PICTURE_PARAMETER_SET_7, pps_val); + } else { + I915_WRITE(ICL_DSC0_PICTURE_PARAMETER_SET_7(pipe), pps_val); + if (crtc_state->dsc_params.dsc_split) + I915_WRITE(ICL_DSC1_PICTURE_PARAMETER_SET_7(pipe), + pps_val); + } + + /* Populate PICTURE_PARAMETER_SET_8 registers */ + pps_val = 0; + pps_val |= DSC_FINAL_OFFSET(vdsc_cfg->final_offset) | + DSC_INITIAL_OFFSET(vdsc_cfg->initial_offset); + DRM_INFO("PPS8 = 0x%08x\n", pps_val); + if (cpu_transcoder == TRANSCODER_EDP) { + I915_WRITE(DSCA_PICTURE_PARAMETER_SET_8, pps_val); + /* + * If 2 VDSC instances are needed, configure PPS for second + * VDSC + */ + if (crtc_state->dsc_params.dsc_split) + I915_WRITE(DSCC_PICTURE_PARAMETER_SET_8, pps_val); + } else { + I915_WRITE(ICL_DSC0_PICTURE_PARAMETER_SET_8(pipe), pps_val); + if (crtc_state->dsc_params.dsc_split) + I915_WRITE(ICL_DSC1_PICTURE_PARAMETER_SET_8(pipe), + pps_val); + } + + /* Populate PICTURE_PARAMETER_SET_9 registers */ + pps_val = 0; + pps_val |= DSC_RC_MODEL_SIZE(DSC_RC_MODEL_SIZE_CONST) | + DSC_RC_EDGE_FACTOR(DSC_RC_EDGE_FACTOR_CONST); + DRM_INFO("PPS9 = 0x%08x\n", pps_val); + if (cpu_transcoder == TRANSCODER_EDP) { + I915_WRITE(DSCA_PICTURE_PARAMETER_SET_9, pps_val); + /* + * If 2 VDSC instances are needed, configure PPS for second + * VDSC + */ + if (crtc_state->dsc_params.dsc_split) + I915_WRITE(DSCC_PICTURE_PARAMETER_SET_9, pps_val); + } else { + I915_WRITE(ICL_DSC0_PICTURE_PARAMETER_SET_9(pipe), pps_val); + if (crtc_state->dsc_params.dsc_split) + I915_WRITE(ICL_DSC1_PICTURE_PARAMETER_SET_9(pipe), + pps_val); + } + + /* Populate PICTURE_PARAMETER_SET_10 registers */ + pps_val = 0; + pps_val |= DSC_RC_QUANT_INC_LIMIT0(vdsc_cfg->rc_quant_incr_limit0) | + DSC_RC_QUANT_INC_LIMIT1(vdsc_cfg->rc_quant_incr_limit1) | + DSC_RC_TARGET_OFF_HIGH(DSC_RC_TGT_OFFSET_HI_CONST) | + DSC_RC_TARGET_OFF_LOW(DSC_RC_TGT_OFFSET_LO_CONST); + DRM_INFO("PPS10 = 0x%08x\n", pps_val); + if (cpu_transcoder == TRANSCODER_EDP) { + I915_WRITE(DSCA_PICTURE_PARAMETER_SET_10, pps_val); + /* + * If 2 VDSC instances are needed, configure PPS for second + * VDSC + */ + if (crtc_state->dsc_params.dsc_split) + I915_WRITE(DSCC_PICTURE_PARAMETER_SET_10, pps_val); + } else { + I915_WRITE(ICL_DSC0_PICTURE_PARAMETER_SET_10(pipe), pps_val); + if (crtc_state->dsc_params.dsc_split) + I915_WRITE(ICL_DSC1_PICTURE_PARAMETER_SET_10(pipe), + pps_val); + } + + /* Populate Picture parameter set 16 */ + pps_val = 0; + pps_val |= DSC_SLICE_CHUNK_SIZE(vdsc_cfg->slice_chunk_size) | + DSC_SLICE_PER_LINE((vdsc_cfg->pic_width / num_vdsc_instances) / + vdsc_cfg->slice_width) | + DSC_SLICE_ROW_PER_FRAME(vdsc_cfg->pic_height / + vdsc_cfg->slice_height); + DRM_INFO("PPS16 = 0x%08x\n", pps_val); + if (cpu_transcoder == TRANSCODER_EDP) { + I915_WRITE(DSCA_PICTURE_PARAMETER_SET_16, pps_val); + /* + * If 2 VDSC instances are needed, configure PPS for second + * VDSC + */ + if (crtc_state->dsc_params.dsc_split) + I915_WRITE(DSCC_PICTURE_PARAMETER_SET_16, pps_val); + } else { + I915_WRITE(ICL_DSC0_PICTURE_PARAMETER_SET_16(pipe), pps_val); + if (crtc_state->dsc_params.dsc_split) + I915_WRITE(ICL_DSC1_PICTURE_PARAMETER_SET_16(pipe), + pps_val); + } + + /* Populate the RC_BUF_THRESH registers */ + memset(rc_buf_thresh_dword, 0, sizeof(rc_buf_thresh_dword)); + for (i = 0; i < DSC_NUM_BUF_RANGES - 1; i++) { + rc_buf_thresh_dword[i / 4] |= + (u32)(vdsc_cfg->rc_buf_thresh[i] << + BITS_PER_BYTE * (i % 4)); + DRM_INFO(" RC_BUF_THRESH%d = 0x%08x\n", i, + rc_buf_thresh_dword[i / 4]); + } + if (cpu_transcoder == TRANSCODER_EDP) { + I915_WRITE(DSCA_RC_BUF_THRESH_0, rc_buf_thresh_dword[0]); + I915_WRITE(DSCA_RC_BUF_THRESH_0_UDW, rc_buf_thresh_dword[1]); + I915_WRITE(DSCA_RC_BUF_THRESH_1, rc_buf_thresh_dword[2]); + I915_WRITE(DSCA_RC_BUF_THRESH_1_UDW, rc_buf_thresh_dword[3]); + if (crtc_state->dsc_params.dsc_split) { + I915_WRITE(DSCC_RC_BUF_THRESH_0, + rc_buf_thresh_dword[0]); + I915_WRITE(DSCC_RC_BUF_THRESH_0_UDW, + rc_buf_thresh_dword[1]); + I915_WRITE(DSCC_RC_BUF_THRESH_1, + rc_buf_thresh_dword[2]); + I915_WRITE(DSCC_RC_BUF_THRESH_1_UDW, + rc_buf_thresh_dword[3]); + } + } else { + I915_WRITE(ICL_DSC0_RC_BUF_THRESH_0(pipe), + rc_buf_thresh_dword[0]); + I915_WRITE(ICL_DSC0_RC_BUF_THRESH_0_UDW(pipe), + rc_buf_thresh_dword[1]); + I915_WRITE(ICL_DSC0_RC_BUF_THRESH_1(pipe), + rc_buf_thresh_dword[2]); + I915_WRITE(ICL_DSC0_RC_BUF_THRESH_1_UDW(pipe), + rc_buf_thresh_dword[3]); + if (crtc_state->dsc_params.dsc_split) { + I915_WRITE(ICL_DSC1_RC_BUF_THRESH_0(pipe), + rc_buf_thresh_dword[0]); + I915_WRITE(ICL_DSC1_RC_BUF_THRESH_0_UDW(pipe), + rc_buf_thresh_dword[1]); + I915_WRITE(ICL_DSC1_RC_BUF_THRESH_1(pipe), + rc_buf_thresh_dword[2]); + I915_WRITE(ICL_DSC1_RC_BUF_THRESH_1_UDW(pipe), + rc_buf_thresh_dword[3]); + } + } + + /* Populate the RC_RANGE_PARAMETERS registers */ + memset(rc_range_params_dword, 0, sizeof(rc_range_params_dword)); + for (i = 0; i < DSC_NUM_BUF_RANGES; i++) { + rc_range_params_dword[i / 2] |= + (u32)(((vdsc_cfg->rc_range_params[i].range_bpg_offset << + RC_BPG_OFFSET_SHIFT) | + (vdsc_cfg->rc_range_params[i].range_max_qp << + RC_MAX_QP_SHIFT) | + (vdsc_cfg->rc_range_params[i].range_min_qp << + RC_MIN_QP_SHIFT)) << 16 * (i % 2)); + DRM_INFO(" RC_RANGE_PARAM_%d = 0x%08x\n", i, + rc_range_params_dword[i / 2]); + } + if (cpu_transcoder == TRANSCODER_EDP) { + I915_WRITE(DSCA_RC_RANGE_PARAMETERS_0, + rc_range_params_dword[0]); + I915_WRITE(DSCA_RC_RANGE_PARAMETERS_0_UDW, + rc_range_params_dword[1]); + I915_WRITE(DSCA_RC_RANGE_PARAMETERS_1, + rc_range_params_dword[2]); + I915_WRITE(DSCA_RC_RANGE_PARAMETERS_1_UDW, + rc_range_params_dword[3]); + I915_WRITE(DSCA_RC_RANGE_PARAMETERS_2, + rc_range_params_dword[4]); + I915_WRITE(DSCA_RC_RANGE_PARAMETERS_2_UDW, + rc_range_params_dword[5]); + I915_WRITE(DSCA_RC_RANGE_PARAMETERS_3, + rc_range_params_dword[6]); + I915_WRITE(DSCA_RC_RANGE_PARAMETERS_3_UDW, + rc_range_params_dword[7]); + if (crtc_state->dsc_params.dsc_split) { + I915_WRITE(DSCC_RC_RANGE_PARAMETERS_0, + rc_range_params_dword[0]); + I915_WRITE(DSCC_RC_RANGE_PARAMETERS_0_UDW, + rc_range_params_dword[1]); + I915_WRITE(DSCC_RC_RANGE_PARAMETERS_1, + rc_range_params_dword[2]); + I915_WRITE(DSCC_RC_RANGE_PARAMETERS_1_UDW, + rc_range_params_dword[3]); + I915_WRITE(DSCC_RC_RANGE_PARAMETERS_2, + rc_range_params_dword[4]); + I915_WRITE(DSCC_RC_RANGE_PARAMETERS_2_UDW, + rc_range_params_dword[5]); + I915_WRITE(DSCC_RC_RANGE_PARAMETERS_3, + rc_range_params_dword[6]); + I915_WRITE(DSCC_RC_RANGE_PARAMETERS_3_UDW, + rc_range_params_dword[7]); + } + } else { + I915_WRITE(ICL_DSC0_RC_RANGE_PARAMETERS_0(pipe), + rc_range_params_dword[0]); + I915_WRITE(ICL_DSC0_RC_RANGE_PARAMETERS_0_UDW(pipe), + rc_range_params_dword[1]); + I915_WRITE(ICL_DSC0_RC_RANGE_PARAMETERS_1(pipe), + rc_range_params_dword[2]); + I915_WRITE(ICL_DSC0_RC_RANGE_PARAMETERS_1_UDW(pipe), + rc_range_params_dword[3]); + I915_WRITE(ICL_DSC0_RC_RANGE_PARAMETERS_2(pipe), + rc_range_params_dword[4]); + I915_WRITE(ICL_DSC0_RC_RANGE_PARAMETERS_2_UDW(pipe), + rc_range_params_dword[5]); + I915_WRITE(ICL_DSC0_RC_RANGE_PARAMETERS_3(pipe), + rc_range_params_dword[6]); + I915_WRITE(ICL_DSC0_RC_RANGE_PARAMETERS_3_UDW(pipe), + rc_range_params_dword[7]); + if (crtc_state->dsc_params.dsc_split) { + I915_WRITE(ICL_DSC1_RC_RANGE_PARAMETERS_0(pipe), + rc_range_params_dword[0]); + I915_WRITE(ICL_DSC1_RC_RANGE_PARAMETERS_0_UDW(pipe), + rc_range_params_dword[1]); + I915_WRITE(ICL_DSC1_RC_RANGE_PARAMETERS_1(pipe), + rc_range_params_dword[2]); + I915_WRITE(ICL_DSC1_RC_RANGE_PARAMETERS_1_UDW(pipe), + rc_range_params_dword[3]); + I915_WRITE(ICL_DSC1_RC_RANGE_PARAMETERS_2(pipe), + rc_range_params_dword[4]); + I915_WRITE(ICL_DSC1_RC_RANGE_PARAMETERS_2_UDW(pipe), + rc_range_params_dword[5]); + I915_WRITE(ICL_DSC1_RC_RANGE_PARAMETERS_3(pipe), + rc_range_params_dword[6]); + I915_WRITE(ICL_DSC1_RC_RANGE_PARAMETERS_3_UDW(pipe), + rc_range_params_dword[7]); + } + } +} + +void intel_dsc_enable(struct intel_encoder *encoder, + const struct intel_crtc_state *crtc_state) +{ + if (!crtc_state->dsc_params.compression_enable) + return; + + intel_configure_pps_for_dsc_encoder(encoder, crtc_state); +} From 4c614831d59bb3dfb3b5516c8dafee0363e4bdf0 Mon Sep 17 00:00:00 2001 From: Manasi Navare Date: Wed, 28 Nov 2018 12:26:20 -0800 Subject: [PATCH 42/98] drm/i915/dp: Use the existing write_infoframe() for DSC PPS SDPs Infoframes are used to send secondary data packets. This patch adds support for DSC Picture parameter set secondary data packets in the existing write_infoframe helpers. v3: * Unused variables cleanup (Ville) v2: * Rebase on drm-tip (Manasi) Cc: Jani Nikula Cc: Ville Syrjala Cc: Anusha Srivatsa Signed-off-by: Manasi Navare Reviewed-by: Anusha Srivatsa Link: https://patchwork.freedesktop.org/patch/msgid/20181128202628.20238-9-manasi.d.navare@intel.com --- drivers/gpu/drm/i915/i915_reg.h | 1 + drivers/gpu/drm/i915/intel_hdmi.c | 21 +++++++++++++++++++-- 2 files changed, 20 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index 47baf2fe8f71..eadd880b0ef6 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -4570,6 +4570,7 @@ enum { * of the infoframe structure specified by CEA-861. */ #define VIDEO_DIP_DATA_SIZE 32 #define VIDEO_DIP_VSC_DATA_SIZE 36 +#define VIDEO_DIP_PPS_DATA_SIZE 132 #define VIDEO_DIP_CTL _MMIO(0x61170) /* Pre HSW: */ #define VIDEO_DIP_ENABLE (1 << 31) diff --git a/drivers/gpu/drm/i915/intel_hdmi.c b/drivers/gpu/drm/i915/intel_hdmi.c index e2c6a2b3e8f2..07e803a604bd 100644 --- a/drivers/gpu/drm/i915/intel_hdmi.c +++ b/drivers/gpu/drm/i915/intel_hdmi.c @@ -115,6 +115,8 @@ static u32 hsw_infoframe_enable(unsigned int type) switch (type) { case DP_SDP_VSC: return VIDEO_DIP_ENABLE_VSC_HSW; + case DP_SDP_PPS: + return VDIP_ENABLE_PPS; case HDMI_INFOFRAME_TYPE_AVI: return VIDEO_DIP_ENABLE_AVI_HSW; case HDMI_INFOFRAME_TYPE_SPD: @@ -136,6 +138,8 @@ hsw_dip_data_reg(struct drm_i915_private *dev_priv, switch (type) { case DP_SDP_VSC: return HSW_TVIDEO_DIP_VSC_DATA(cpu_transcoder, i); + case DP_SDP_PPS: + return ICL_VIDEO_DIP_PPS_DATA(cpu_transcoder, i); case HDMI_INFOFRAME_TYPE_AVI: return HSW_TVIDEO_DIP_AVI_DATA(cpu_transcoder, i); case HDMI_INFOFRAME_TYPE_SPD: @@ -148,6 +152,18 @@ hsw_dip_data_reg(struct drm_i915_private *dev_priv, } } +static int hsw_dip_data_size(unsigned int type) +{ + switch (type) { + case DP_SDP_VSC: + return VIDEO_DIP_VSC_DATA_SIZE; + case DP_SDP_PPS: + return VIDEO_DIP_PPS_DATA_SIZE; + default: + return VIDEO_DIP_DATA_SIZE; + } +} + static void g4x_write_infoframe(struct intel_encoder *encoder, const struct intel_crtc_state *crtc_state, unsigned int type, @@ -382,11 +398,12 @@ static void hsw_write_infoframe(struct intel_encoder *encoder, struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); enum transcoder cpu_transcoder = crtc_state->cpu_transcoder; i915_reg_t ctl_reg = HSW_TVIDEO_DIP_CTL(cpu_transcoder); - int data_size = type == DP_SDP_VSC ? - VIDEO_DIP_VSC_DATA_SIZE : VIDEO_DIP_DATA_SIZE; + int data_size; int i; u32 val = I915_READ(ctl_reg); + data_size = hsw_dip_data_size(type); + val &= ~hsw_infoframe_enable(type); I915_WRITE(ctl_reg, val); From 5b1ea77228f913df20a445512bd57ea481905d4e Mon Sep 17 00:00:00 2001 From: Manasi Navare Date: Wed, 28 Nov 2018 12:26:21 -0800 Subject: [PATCH 43/98] drm/i915/dp: Populate DSC PPS SDP and send PPS infoframes DSC PPS secondary data packet infoframes are filled with DSC picure parameter set metadata according to the DSC standard. These infoframes are sent to the sink device and used during DSC decoding. v3: * Rename to intel_dp_write_pps_sdp (Ville) * Use const intel_crtc_state (Ville) v2: * Rebase ond drm-tip Cc: Jani Nikula Cc: Ville Syrjala Cc: Anusha Srivatsa Signed-off-by: Manasi Navare Reviewed-by: Anusha Srivatsa Link: https://patchwork.freedesktop.org/patch/msgid/20181128202628.20238-10-manasi.d.navare@intel.com --- drivers/gpu/drm/i915/intel_vdsc.c | 21 +++++++++++++++++++++ 1 file changed, 21 insertions(+) diff --git a/drivers/gpu/drm/i915/intel_vdsc.c b/drivers/gpu/drm/i915/intel_vdsc.c index 9676945a85bb..0b31eac46988 100644 --- a/drivers/gpu/drm/i915/intel_vdsc.c +++ b/drivers/gpu/drm/i915/intel_vdsc.c @@ -978,6 +978,25 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder, } } +static void intel_dp_write_dsc_pps_sdp(struct intel_encoder *encoder, + const struct intel_crtc_state *crtc_state) +{ + struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base); + struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp); + const struct drm_dsc_config *vdsc_cfg = &crtc_state->dp_dsc_cfg; + struct drm_dsc_pps_infoframe dp_dsc_pps_sdp; + + /* Prepare DP SDP PPS header as per DP 1.4 spec, Table 2-123 */ + drm_dsc_dp_pps_header_init(&dp_dsc_pps_sdp); + + /* Fill the PPS payload bytes as per DSC spec 1.2 Table 4-1 */ + drm_dsc_pps_infoframe_pack(&dp_dsc_pps_sdp, vdsc_cfg); + + intel_dig_port->write_infoframe(encoder, crtc_state, + DP_SDP_PPS, &dp_dsc_pps_sdp, + sizeof(dp_dsc_pps_sdp)); +} + void intel_dsc_enable(struct intel_encoder *encoder, const struct intel_crtc_state *crtc_state) { @@ -985,4 +1004,6 @@ void intel_dsc_enable(struct intel_encoder *encoder, return; intel_configure_pps_for_dsc_encoder(encoder, crtc_state); + + intel_dp_write_dsc_pps_sdp(encoder, crtc_state); } From a311b0b5d2094029dce2369d686044131e19e006 Mon Sep 17 00:00:00 2001 From: Manasi Navare Date: Wed, 28 Nov 2018 12:26:22 -0800 Subject: [PATCH 44/98] drm/i915/dp: Configure Display stream splitter registers during DSC enable Display Stream Splitter registers need to be programmed to enable the joiner if two DSC engines are used and also to enable the left and the right DSC engines. This happens as part of the DSC enabling routine in the source in atomic commit. v4: * Remove redundant comment (Ville) v3: * Use cpu_transcoder instead of encoder->type (Ville) v2: * Rebase (Manasi) Cc: Jani Nikula Cc: Ville Syrjala Cc: Anusha Srivatsa Signed-off-by: Manasi Navare Reviewed-by: Anusha Srivatsa Link: https://patchwork.freedesktop.org/patch/msgid/20181128202628.20238-11-manasi.d.navare@intel.com --- drivers/gpu/drm/i915/intel_vdsc.c | 22 ++++++++++++++++++++++ 1 file changed, 22 insertions(+) diff --git a/drivers/gpu/drm/i915/intel_vdsc.c b/drivers/gpu/drm/i915/intel_vdsc.c index 0b31eac46988..696cb4b8c9ea 100644 --- a/drivers/gpu/drm/i915/intel_vdsc.c +++ b/drivers/gpu/drm/i915/intel_vdsc.c @@ -1000,10 +1000,32 @@ static void intel_dp_write_dsc_pps_sdp(struct intel_encoder *encoder, void intel_dsc_enable(struct intel_encoder *encoder, const struct intel_crtc_state *crtc_state) { + struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc); + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); + enum pipe pipe = crtc->pipe; + i915_reg_t dss_ctl1_reg, dss_ctl2_reg; + u32 dss_ctl1_val = 0; + u32 dss_ctl2_val = 0; + if (!crtc_state->dsc_params.compression_enable) return; intel_configure_pps_for_dsc_encoder(encoder, crtc_state); intel_dp_write_dsc_pps_sdp(encoder, crtc_state); + + if (crtc_state->cpu_transcoder == TRANSCODER_EDP) { + dss_ctl1_reg = DSS_CTL1; + dss_ctl2_reg = DSS_CTL2; + } else { + dss_ctl1_reg = ICL_PIPE_DSS_CTL1(pipe); + dss_ctl2_reg = ICL_PIPE_DSS_CTL2(pipe); + } + dss_ctl2_val |= LEFT_BRANCH_VDSC_ENABLE; + if (crtc_state->dsc_params.dsc_split) { + dss_ctl2_val |= RIGHT_BRANCH_VDSC_ENABLE; + dss_ctl1_val |= JOINER_ENABLE; + } + I915_WRITE(dss_ctl1_reg, dss_ctl1_val); + I915_WRITE(dss_ctl2_reg, dss_ctl2_val); } From a600622c09ddf7da660ca714d5644ecf270426fc Mon Sep 17 00:00:00 2001 From: Manasi Navare Date: Wed, 28 Nov 2018 12:26:23 -0800 Subject: [PATCH 45/98] drm/i915/dp: Disable DSC in source by disabling DSS CTL bits 1. Disable Left/right VDSC branch in DSS Ctrl reg depending on the number of VDSC engines being used 2. Disable joiner in DSS Ctrl reg v4: * Remove encoder, make crtc_state const (Ville) v3 (From Manasi): * Add Disable PG2 for VDSC on eDP v2 (From Manasi): * Use old_crtc_state to find dsc params * Add a condition to disable only if dsc state compression is enabled * Use correct DSS CTL regs Cc: Jani Nikula Cc: Ville Syrjala Cc: Anusha Srivatsa Signed-off-by: Manasi Navare Signed-off-by: Gaurav K Singh Reviewed-by: Anusha Srivatsa Link: https://patchwork.freedesktop.org/patch/msgid/20181128202628.20238-12-manasi.d.navare@intel.com --- drivers/gpu/drm/i915/i915_drv.h | 1 + drivers/gpu/drm/i915/intel_display.c | 2 ++ drivers/gpu/drm/i915/intel_vdsc.c | 31 ++++++++++++++++++++++++++++ 3 files changed, 34 insertions(+) diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 06d1eaeeba7a..43ac6873a2bb 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -3342,6 +3342,7 @@ extern bool intel_set_memory_cxsr(struct drm_i915_private *dev_priv, bool enable); void intel_dsc_enable(struct intel_encoder *encoder, const struct intel_crtc_state *crtc_state); +void intel_dsc_disable(const struct intel_crtc_state *crtc_state); int i915_reg_read_ioctl(struct drm_device *dev, void *data, struct drm_file *file); diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index f3ba04562cbd..789f647bd598 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -5923,6 +5923,8 @@ static void haswell_crtc_disable(struct intel_crtc_state *old_crtc_state, if (!transcoder_is_dsi(cpu_transcoder)) intel_ddi_disable_transcoder_func(old_crtc_state); + intel_dsc_disable(old_crtc_state); + if (INTEL_GEN(dev_priv) >= 9) skylake_scaler_disable(intel_crtc); else diff --git a/drivers/gpu/drm/i915/intel_vdsc.c b/drivers/gpu/drm/i915/intel_vdsc.c index 696cb4b8c9ea..ec7444f0ca12 100644 --- a/drivers/gpu/drm/i915/intel_vdsc.c +++ b/drivers/gpu/drm/i915/intel_vdsc.c @@ -1029,3 +1029,34 @@ void intel_dsc_enable(struct intel_encoder *encoder, I915_WRITE(dss_ctl1_reg, dss_ctl1_val); I915_WRITE(dss_ctl2_reg, dss_ctl2_val); } + +void intel_dsc_disable(const struct intel_crtc_state *old_crtc_state) +{ + struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->base.crtc); + struct drm_i915_private *dev_priv = to_i915(crtc->base.dev); + enum pipe pipe = crtc->pipe; + i915_reg_t dss_ctl1_reg, dss_ctl2_reg; + u32 dss_ctl1_val = 0, dss_ctl2_val = 0; + + if (!old_crtc_state->dsc_params.compression_enable) + return; + + if (old_crtc_state->cpu_transcoder == TRANSCODER_EDP) { + dss_ctl1_reg = DSS_CTL1; + dss_ctl2_reg = DSS_CTL2; + } else { + dss_ctl1_reg = ICL_PIPE_DSS_CTL1(pipe); + dss_ctl2_reg = ICL_PIPE_DSS_CTL2(pipe); + } + dss_ctl1_val = I915_READ(dss_ctl1_reg); + if (dss_ctl1_val & JOINER_ENABLE) + dss_ctl1_val &= ~JOINER_ENABLE; + I915_WRITE(dss_ctl1_reg, dss_ctl1_val); + + dss_ctl2_val = I915_READ(dss_ctl2_reg); + if (dss_ctl2_val & LEFT_BRANCH_VDSC_ENABLE || + dss_ctl2_val & RIGHT_BRANCH_VDSC_ENABLE) + dss_ctl2_val &= ~(LEFT_BRANCH_VDSC_ENABLE | + RIGHT_BRANCH_VDSC_ENABLE); + I915_WRITE(dss_ctl2_reg, dss_ctl2_val); +} From a24c62f94be1a478ceb1086df726615a73a9e113 Mon Sep 17 00:00:00 2001 From: Manasi Navare Date: Wed, 28 Nov 2018 12:26:24 -0800 Subject: [PATCH 46/98] drm/i915/dsc: Enable and disable appropriate power wells for VDSC MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit A separate power well 2 (PG2) is required for VDSC on eDP transcoder whereas all other transcoders use the power wells associated with the transcoders for VDSC. This patch adds a helper to obtain correct power domain depending on transcoder being used and enables/disables the power wells during VDSC enabling/disabling. v4: * Get VDSC power domain only if compression en is set in crtc_state (Ville, Imre) v3: * Call it intel_dsc_power_domain, add to intel_ddi_get_power_domains (Ville) v2: * Fix tabs, const crtc_state, fix comments (Ville) Suggested-by: Ville Syrjala Cc: Ville Syrjala Cc: Imre Deak Cc: Rodrigo Vivi Signed-off-by: Manasi Navare Reviewed-by: Ville Syrjälä Link: https://patchwork.freedesktop.org/patch/msgid/20181128202628.20238-13-manasi.d.navare@intel.com --- drivers/gpu/drm/i915/intel_ddi.c | 6 ++++++ drivers/gpu/drm/i915/intel_drv.h | 2 ++ drivers/gpu/drm/i915/intel_vdsc.c | 26 ++++++++++++++++++++++++++ 3 files changed, 34 insertions(+) diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c index 339be10986d7..6533624226a7 100644 --- a/drivers/gpu/drm/i915/intel_ddi.c +++ b/drivers/gpu/drm/i915/intel_ddi.c @@ -2154,6 +2154,12 @@ static u64 intel_ddi_get_power_domains(struct intel_encoder *encoder, intel_port_is_tc(dev_priv, encoder->port)) domains |= BIT_ULL(intel_ddi_main_link_aux_domain(dig_port)); + /* + * VDSC power is needed when DSC is enabled + */ + if (crtc_state->dsc_params.compression_enable) + domains |= BIT_ULL(intel_dsc_power_domain(crtc_state)); + return domains; } diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index 36e756e15124..0623dc8ab115 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -1863,6 +1863,8 @@ uint8_t intel_dp_dsc_get_slice_count(struct intel_dp *intel_dp, int mode_clock, /* intel_vdsc.c */ int intel_dp_compute_dsc_params(struct intel_dp *intel_dp, struct intel_crtc_state *pipe_config); +enum intel_display_power_domain +intel_dsc_power_domain(const struct intel_crtc_state *crtc_state); static inline unsigned int intel_dp_unused_lane_mask(int lane_count) { diff --git a/drivers/gpu/drm/i915/intel_vdsc.c b/drivers/gpu/drm/i915/intel_vdsc.c index ec7444f0ca12..c56ba0e04044 100644 --- a/drivers/gpu/drm/i915/intel_vdsc.c +++ b/drivers/gpu/drm/i915/intel_vdsc.c @@ -578,6 +578,24 @@ int intel_dp_compute_dsc_params(struct intel_dp *intel_dp, return intel_compute_rc_parameters(vdsc_cfg); } +enum intel_display_power_domain +intel_dsc_power_domain(const struct intel_crtc_state *crtc_state) +{ + enum transcoder cpu_transcoder = crtc_state->cpu_transcoder; + + /* + * On ICL VDSC/joining for eDP transcoder uses a separate power well PW2 + * This requires POWER_DOMAIN_TRANSCODER_EDP_VDSC power domain. + * For any other transcoder, VDSC/joining uses the power well associated + * with the pipe/transcoder in use. Hence another reference on the + * transcoder power domain will suffice. + */ + if (cpu_transcoder == TRANSCODER_EDP) + return POWER_DOMAIN_TRANSCODER_EDP_VDSC; + else + return POWER_DOMAIN_TRANSCODER(cpu_transcoder); +} + static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder, const struct intel_crtc_state *crtc_state) { @@ -1010,6 +1028,10 @@ void intel_dsc_enable(struct intel_encoder *encoder, if (!crtc_state->dsc_params.compression_enable) return; + /* Enable Power wells for VDSC/joining */ + intel_display_power_get(dev_priv, + intel_dsc_power_domain(crtc_state)); + intel_configure_pps_for_dsc_encoder(encoder, crtc_state); intel_dp_write_dsc_pps_sdp(encoder, crtc_state); @@ -1059,4 +1081,8 @@ void intel_dsc_disable(const struct intel_crtc_state *old_crtc_state) dss_ctl2_val &= ~(LEFT_BRANCH_VDSC_ENABLE | RIGHT_BRANCH_VDSC_ENABLE); I915_WRITE(dss_ctl2_reg, dss_ctl2_val); + + /* Disable Power wells for VDSC/joining */ + intel_display_power_put(dev_priv, + intel_dsc_power_domain(old_crtc_state)); } From 240999cf339f146d186dfb1a82ac6ba77ef34a1a Mon Sep 17 00:00:00 2001 From: Anusha Srivatsa Date: Wed, 28 Nov 2018 12:26:25 -0800 Subject: [PATCH 47/98] i915/dp/fec: Add fec_enable to the crtc state. For DP 1.4 and above, Display Stream compression can be enabled only if Forward Error Correctin can be performed. Add a crtc state for FEC. Currently, the state is determined by platform, DP and DSC being enabled. Moving forward we can use the state to have error correction on other scenarios too if needed. v2: - Control compression_enable with the fec_enable parameter in crtc state and with intel_dp_supports_fec() (Ville) - intel_dp_can_fec()/intel_dp_supports_fec()(manasi) v3: Check for FEC support along with setting crtc state. v4: add checks to intel_dp_source_supports_dsc.(manasi) - Move intel_dp_supports_fec() closer to intel_dp_supports_dsc() (Anusha) v5: Move fec check to intel_dp_supports_dsc(Ville) v6: Remove warning. rebase. v7: change crtc state to include DP sink and fec capability of source.(Manasi) v8: Set fec_enable in crtc in intel_dp_compute_config(). v9 (From Manasi): * Combine the !edp and !fec_support check * Derive dev_priv from intel_dp directly v10 (From Manasi): * Rebase Suggested-by: Ville Syrjala Cc: dri-devel@lists.freedesktop.org Cc: Ville Syrjala Cc: Jani Nikula Cc: Manasi Navare Signed-off-by: Anusha Srivatsa Signed-off-by: Manasi Navare Reviewed-by: Manasi Navare Link: https://patchwork.freedesktop.org/patch/msgid/20181128202628.20238-14-manasi.d.navare@intel.com --- drivers/gpu/drm/i915/intel_dp.c | 28 +++++++++++++++++++++++----- drivers/gpu/drm/i915/intel_drv.h | 3 +++ 2 files changed, 26 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 4e65dd1a6c07..38a6e82153fd 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -545,7 +545,7 @@ intel_dp_mode_valid(struct drm_connector *connector, dsc_slice_count = drm_dp_dsc_sink_max_slice_count(intel_dp->dsc_dpcd, true); - } else { + } else if (drm_dp_sink_supports_fec(intel_dp->fec_capable)) { dsc_max_output_bpp = intel_dp_dsc_get_output_bpp(max_link_clock, max_lanes, @@ -1710,14 +1710,26 @@ struct link_config_limits { int min_bpp, max_bpp; }; -static bool intel_dp_source_supports_dsc(struct intel_dp *intel_dp, +static bool intel_dp_source_supports_fec(struct intel_dp *intel_dp, const struct intel_crtc_state *pipe_config) { struct drm_i915_private *dev_priv = dp_to_i915(intel_dp); - /* FIXME: FEC needed for external DP until then reject DSC on DP */ - if (!intel_dp_is_edp(intel_dp)) - return false; + return INTEL_GEN(dev_priv) >= 11 && + pipe_config->cpu_transcoder != TRANSCODER_A; +} + +static bool intel_dp_supports_fec(struct intel_dp *intel_dp, + const struct intel_crtc_state *pipe_config) +{ + return intel_dp_source_supports_fec(intel_dp, pipe_config) && + drm_dp_sink_supports_fec(intel_dp->fec_capable); +} + +static bool intel_dp_source_supports_dsc(struct intel_dp *intel_dp, + const struct intel_crtc_state *pipe_config) +{ + struct drm_i915_private *dev_priv = dp_to_i915(intel_dp); return INTEL_GEN(dev_priv) >= 10 && pipe_config->cpu_transcoder != TRANSCODER_A; @@ -1726,6 +1738,9 @@ static bool intel_dp_source_supports_dsc(struct intel_dp *intel_dp, static bool intel_dp_supports_dsc(struct intel_dp *intel_dp, const struct intel_crtc_state *pipe_config) { + if (!intel_dp_is_edp(intel_dp) && !pipe_config->fec_enable) + return false; + return intel_dp_source_supports_dsc(intel_dp, pipe_config) && drm_dp_sink_supports_dsc(intel_dp->dsc_dpcd); } @@ -2129,6 +2144,9 @@ intel_dp_compute_config(struct intel_encoder *encoder, if (adjusted_mode->flags & DRM_MODE_FLAG_DBLCLK) return false; + pipe_config->fec_enable = !intel_dp_is_edp(intel_dp) && + intel_dp_supports_fec(intel_dp, pipe_config); + if (!intel_dp_compute_link_config(encoder, pipe_config, conn_state)) return false; diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index 0623dc8ab115..86ff57738cd9 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -951,6 +951,9 @@ struct intel_crtc_state { u8 slice_count; } dsc_params; struct drm_dsc_config dp_dsc_cfg; + + /* Forward Error correction State */ + bool fec_enable; }; struct intel_crtc { From a322b97589a6af4b87978d200645aac562fc746f Mon Sep 17 00:00:00 2001 From: Anusha Srivatsa Date: Wed, 28 Nov 2018 12:26:26 -0800 Subject: [PATCH 48/98] drm/i915/fec: Set FEC_READY in FEC_CONFIGURATION If the panel supports FEC, the driver has to set the FEC_READY bit in the dpcd register: FEC_CONFIGURATION. This has to happen before link training. v2: s/intel_dp_set_fec_ready/intel_dp_sink_set_fec_ready - change commit message. (Gaurav) v3: rebased. (r-b Manasi) v4: Use fec crtc state, before setting FEC_READY bit. (Anusha) v5: Move to intel_ddi.c - Make the function static (Anusha) v6: Dont pass state as a separate argument (Ville) v7: (From Manasi) * Correct the debug print (Ville) Cc: dri-devel@lists.freedesktop.org Cc: Gaurav K Singh Cc: Jani Nikula Cc: Ville Syrjala Cc: Manasi Navare Signed-off-by: Anusha Srivatsa Signed-off-by: Manasi Navare Reviewed-by: Manasi Navare Link: https://patchwork.freedesktop.org/patch/msgid/20181128202628.20238-15-manasi.d.navare@intel.com --- drivers/gpu/drm/i915/intel_ddi.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c index 6533624226a7..4c74bbe1cf73 100644 --- a/drivers/gpu/drm/i915/intel_ddi.c +++ b/drivers/gpu/drm/i915/intel_ddi.c @@ -3102,6 +3102,16 @@ static void icl_program_mg_dp_mode(struct intel_digital_port *intel_dig_port) I915_WRITE(MG_DP_MODE(port, 1), ln1); } +static void intel_dp_sink_set_fec_ready(struct intel_dp *intel_dp, + const struct intel_crtc_state *crtc_state) +{ + if (!crtc_state->fec_enable) + return; + + if (drm_dp_dpcd_writeb(&intel_dp->aux, DP_FEC_CONFIGURATION, DP_FEC_READY) <= 0) + DRM_DEBUG_KMS("Failed to set FEC_READY in the sink\n"); +} + static void intel_ddi_pre_enable_dp(struct intel_encoder *encoder, const struct intel_crtc_state *crtc_state, const struct drm_connector_state *conn_state) @@ -3142,6 +3152,7 @@ static void intel_ddi_pre_enable_dp(struct intel_encoder *encoder, intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_ON); intel_dp_sink_set_decompression_state(intel_dp, crtc_state, true); + intel_dp_sink_set_fec_ready(intel_dp, crtc_state); intel_dp_start_link_train(intel_dp); if (port != PORT_A || INTEL_GEN(dev_priv) >= 9) intel_dp_stop_link_train(intel_dp); From 5c44b938629a81641e283f0b84e1d8cf48cbdc41 Mon Sep 17 00:00:00 2001 From: Anusha Srivatsa Date: Wed, 28 Nov 2018 12:26:27 -0800 Subject: [PATCH 49/98] i915/dp/fec: Configure the Forward Error Correction bits. If FEC is supported, the corresponding DP_TP_CTL register bits have to be configured. The driver has to program the FEC_ENABLE in DP_TP_CTL[30] register and wait till FEC_STATUS in DP_TP_CTL[28] is 1. Also add the warn message to make sure that the control register is already active while enabling FEC. v2: - Change commit message. Configure fec state after link training (Manasi, Gaurav) - Remove redundent checks (Manasi) - Remove the registers that get added automagically (Anusha) v3: s/intel_dp_set_fec_state()/intel_dp_enable_fec_state() (Gaurav) v4: rebased. v5: - Move the code to the proper spot, according to spec.(Ville) - Use fec state as a check too. v6: Pass intel_encoder, instead of intel_dp. (Ville) v7: Remove unwanted comments (Manasi) Cc: dri-devel@lists.freedesktop.org Cc: Gaurav K Singh Cc: Jani Nikula Cc: Ville Syrjala Cc: Manasi Navare Signed-off-by: Anusha Srivatsa Signed-off-by: Manasi Navare Reviewed-by: Manasi Navare Link: https://patchwork.freedesktop.org/patch/msgid/20181128202628.20238-16-manasi.d.navare@intel.com --- drivers/gpu/drm/i915/i915_reg.h | 2 ++ drivers/gpu/drm/i915/intel_ddi.c | 23 +++++++++++++++++++++++ 2 files changed, 25 insertions(+) diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index eadd880b0ef6..d3ef97915455 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -9198,6 +9198,7 @@ enum skl_power_gate { #define _DP_TP_CTL_B 0x64140 #define DP_TP_CTL(port) _MMIO_PORT(port, _DP_TP_CTL_A, _DP_TP_CTL_B) #define DP_TP_CTL_ENABLE (1 << 31) +#define DP_TP_CTL_FEC_ENABLE (1 << 30) #define DP_TP_CTL_MODE_SST (0 << 27) #define DP_TP_CTL_MODE_MST (1 << 27) #define DP_TP_CTL_FORCE_ACT (1 << 25) @@ -9216,6 +9217,7 @@ enum skl_power_gate { #define _DP_TP_STATUS_A 0x64044 #define _DP_TP_STATUS_B 0x64144 #define DP_TP_STATUS(port) _MMIO_PORT(port, _DP_TP_STATUS_A, _DP_TP_STATUS_B) +#define DP_TP_STATUS_FEC_ENABLE_LIVE (1 << 28) #define DP_TP_STATUS_IDLE_DONE (1 << 25) #define DP_TP_STATUS_ACT_SENT (1 << 24) #define DP_TP_STATUS_MODE_STATUS_MST (1 << 23) diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c index 4c74bbe1cf73..12acdb08a750 100644 --- a/drivers/gpu/drm/i915/intel_ddi.c +++ b/drivers/gpu/drm/i915/intel_ddi.c @@ -3112,6 +3112,27 @@ static void intel_dp_sink_set_fec_ready(struct intel_dp *intel_dp, DRM_DEBUG_KMS("Failed to set FEC_READY in the sink\n"); } +static void intel_ddi_enable_fec(struct intel_encoder *encoder, + const struct intel_crtc_state *crtc_state) +{ + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); + enum port port = encoder->port; + u32 val; + + if (!crtc_state->fec_enable) + return; + + val = I915_READ(DP_TP_CTL(port)); + val |= DP_TP_CTL_FEC_ENABLE; + I915_WRITE(DP_TP_CTL(port), val); + + if (intel_wait_for_register(dev_priv, DP_TP_STATUS(port), + DP_TP_STATUS_FEC_ENABLE_LIVE, + DP_TP_STATUS_FEC_ENABLE_LIVE, + 1)) + DRM_ERROR("Timed out waiting for FEC Enable Status\n"); +} + static void intel_ddi_pre_enable_dp(struct intel_encoder *encoder, const struct intel_crtc_state *crtc_state, const struct drm_connector_state *conn_state) @@ -3157,6 +3178,8 @@ static void intel_ddi_pre_enable_dp(struct intel_encoder *encoder, if (port != PORT_A || INTEL_GEN(dev_priv) >= 9) intel_dp_stop_link_train(intel_dp); + intel_ddi_enable_fec(encoder, crtc_state); + icl_enable_phy_clock_gating(dig_port); if (!is_mst) From d6a09cee2458adeed5c07882c7e0b91f089515b8 Mon Sep 17 00:00:00 2001 From: Anusha Srivatsa Date: Wed, 28 Nov 2018 12:26:28 -0800 Subject: [PATCH 50/98] drm/i915/fec: Disable FEC state. Set the suitable bits in DP_TP_CTL to stop bit correction when DSC is disabled. v2: - rebased. - Add additional check for compression state. (Gaurav) v3: rebased. v4: - Move the code to the proper spot according to spec (Ville) - Use proper checks (manasi) v5: Remove unnecessary checks (Ville) v6: Resolve warnings. Add crtc_state as an argument to intel_disable_ddi_buf(). (Manasi) Cc: dri-devel@lists.freedesktop.org Cc: Gaurav K Singh Cc: Jani Nikula Cc: Ville Syrjala Cc: Manasi Navare Signed-off-by: Anusha Srivatsa Signed-off-by: Manasi Navare Reviewed-by: Manasi Navare Link: https://patchwork.freedesktop.org/patch/msgid/20181128202628.20238-17-manasi.d.navare@intel.com --- drivers/gpu/drm/i915/intel_ddi.c | 28 ++++++++++++++++++++++++---- 1 file changed, 24 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c index 12acdb08a750..61d7145f93bf 100644 --- a/drivers/gpu/drm/i915/intel_ddi.c +++ b/drivers/gpu/drm/i915/intel_ddi.c @@ -3133,6 +3133,22 @@ static void intel_ddi_enable_fec(struct intel_encoder *encoder, DRM_ERROR("Timed out waiting for FEC Enable Status\n"); } +static void intel_ddi_disable_fec_state(struct intel_encoder *encoder, + const struct intel_crtc_state *crtc_state) +{ + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); + enum port port = encoder->port; + u32 val; + + if (!crtc_state->fec_enable) + return; + + val = I915_READ(DP_TP_CTL(port)); + val &= ~DP_TP_CTL_FEC_ENABLE; + I915_WRITE(DP_TP_CTL(port), val); + POSTING_READ(DP_TP_CTL(port)); +} + static void intel_ddi_pre_enable_dp(struct intel_encoder *encoder, const struct intel_crtc_state *crtc_state, const struct drm_connector_state *conn_state) @@ -3272,7 +3288,8 @@ static void intel_ddi_pre_enable(struct intel_encoder *encoder, } } -static void intel_disable_ddi_buf(struct intel_encoder *encoder) +static void intel_disable_ddi_buf(struct intel_encoder *encoder, + const struct intel_crtc_state *crtc_state) { struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); enum port port = encoder->port; @@ -3291,6 +3308,9 @@ static void intel_disable_ddi_buf(struct intel_encoder *encoder) val |= DP_TP_CTL_LINK_TRAIN_PAT1; I915_WRITE(DP_TP_CTL(port), val); + /* Disable FEC in DP Sink */ + intel_ddi_disable_fec_state(encoder, crtc_state); + if (wait) intel_wait_ddi_buf_idle(dev_priv, port); } @@ -3314,7 +3334,7 @@ static void intel_ddi_post_disable_dp(struct intel_encoder *encoder, intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_OFF); } - intel_disable_ddi_buf(encoder); + intel_disable_ddi_buf(encoder, old_crtc_state); intel_edp_panel_vdd_on(intel_dp); intel_edp_panel_off(intel_dp); @@ -3337,7 +3357,7 @@ static void intel_ddi_post_disable_hdmi(struct intel_encoder *encoder, intel_ddi_disable_pipe_clock(old_crtc_state); - intel_disable_ddi_buf(encoder); + intel_disable_ddi_buf(encoder, old_crtc_state); intel_display_power_put(dev_priv, dig_port->ddi_io_power_domain); @@ -3388,7 +3408,7 @@ void intel_ddi_fdi_post_disable(struct intel_encoder *encoder, val &= ~FDI_RX_ENABLE; I915_WRITE(FDI_RX_CTL(PIPE_A), val); - intel_disable_ddi_buf(encoder); + intel_disable_ddi_buf(encoder, old_crtc_state); intel_ddi_clk_disable(encoder); val = I915_READ(FDI_RX_MISC(PIPE_A)); From 5a756aadf5f7472422bbbfa373cdf090bcb343ec Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Thu, 29 Nov 2018 21:05:43 +0000 Subject: [PATCH 51/98] drm/i915: Fixup stub definitions for intel_opregion_suspend|resume 248 "multiple definition of ...". E.g.: LD [M] drivers/gpu/drm/i915/i915.o ld: drivers/gpu/drm/i915/i915_irq.o: in function `intel_opregion_resume': i915_irq.c:(.text+0x58f0): multiple definition of `intel_opregion_resume'; drivers/gpu/drm/i915/i915_drv.o:i915_drv.c:(.text+0x2d40): first defined here Reported-by: Randy Dunlap Reported-by: Stephen Rothwell Fixes: a950adc6c343 ("drm/i915: Stop calling intel_opregion unregister/register in suspend/resume") Signed-off-by: Chris Wilson Acked-by: Randy Dunlap # build tested Link: https://patchwork.freedesktop.org/patch/msgid/20181129210543.31555-1-chris@chris-wilson.co.uk --- drivers/gpu/drm/i915/intel_opregion.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_opregion.h b/drivers/gpu/drm/i915/intel_opregion.h index d84b6d2d2fae..4aa68ffbd30e 100644 --- a/drivers/gpu/drm/i915/intel_opregion.h +++ b/drivers/gpu/drm/i915/intel_opregion.h @@ -87,12 +87,12 @@ static inline void intel_opregion_unregister(struct drm_i915_private *dev_priv) { } -void intel_opregion_resume(struct drm_i915_private *dev_priv) +static inline void intel_opregion_resume(struct drm_i915_private *dev_priv) { } -void intel_opregion_suspend(struct drm_i915_private *dev_priv, - pci_power_t state) +static inline void intel_opregion_suspend(struct drm_i915_private *dev_priv, + pci_power_t state) { } From a2538cbc989cfab007149dddd1b63a0ba509370a Mon Sep 17 00:00:00 2001 From: Tvrtko Ursulin Date: Thu, 29 Nov 2018 13:41:28 +0000 Subject: [PATCH 52/98] drm/i915: Remove whitelist application from ringbuffer backend There is no white-listing before Gen8 and after the removal ringbuffer support for these platforms we can remove the call to this no-op. Signed-off-by: Tvrtko Ursulin Reviewed-by: Chris Wilson Link: https://patchwork.freedesktop.org/patch/msgid/20181129134128.7994-1-tvrtko.ursulin@linux.intel.com --- drivers/gpu/drm/i915/intel_ringbuffer.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index e18a64d41843..28ae1e436ea6 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -667,8 +667,6 @@ static int init_render_ring(struct intel_engine_cs *engine) if (ret) return ret; - intel_whitelist_workarounds_apply(engine); - /* WaTimedSingleVertexDispatch:cl,bw,ctg,elk,ilk,snb */ if (IS_GEN(dev_priv, 4, 6)) I915_WRITE(MI_MODE, _MASKED_BIT_ENABLE(VS_TIMER_DISPATCH)); From f545425a0145dc6cc9c6aca0f69d0bf7ca198906 Mon Sep 17 00:00:00 2001 From: Tvrtko Ursulin Date: Thu, 29 Nov 2018 13:46:30 +0000 Subject: [PATCH 53/98] drm/i915/icl: Remove Wa_1604302699 It seems that the documentation does not reference this any more, plus, bit 28 does not stick when written to the register. Therefore I can only assume this is something which was documented in the past but got removed from the hardware in the meantime. Signed-off-by: Tvrtko Ursulin Cc: Mika Kuoppala Reviewed-by: Mika Kuoppala Link: https://patchwork.freedesktop.org/patch/msgid/20181129134630.8222-1-tvrtko.ursulin@linux.intel.com --- drivers/gpu/drm/i915/intel_workarounds.c | 5 ----- 1 file changed, 5 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c index 39cd1f823ea9..e5cd6c6c66c3 100644 --- a/drivers/gpu/drm/i915/intel_workarounds.c +++ b/drivers/gpu/drm/i915/intel_workarounds.c @@ -882,11 +882,6 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv) I915_READ(SUBSLICE_UNIT_LEVEL_CLKGATE) | GWUNIT_CLKGATE_DIS); - /* Wa_1604302699:icl */ - I915_WRITE(GEN10_L3_CHICKEN_MODE_REGISTER, - I915_READ(GEN10_L3_CHICKEN_MODE_REGISTER) | - GEN11_I2M_WRITE_DISABLE); - /* Wa_1406838659:icl (pre-prod) */ if (IS_ICL_REVID(dev_priv, ICL_REVID_A0, ICL_REVID_B0)) I915_WRITE(INF_UNIT_LEVEL_CLKGATE, From 8d2f6e2f2721097549dded41aa6e1c1a4e8e12ee Mon Sep 17 00:00:00 2001 From: Tvrtko Ursulin Date: Fri, 30 Nov 2018 08:02:53 +0000 Subject: [PATCH 54/98] drm/i915/selftests: Extract spinner code Pull out spinner code to a standalone file to enable it to be shortly used by other and new test cases. Plain code movement - no functional changes. Signed-off-by: Tvrtko Ursulin Reviewed-by: Chris Wilson Link: https://patchwork.freedesktop.org/patch/msgid/20181130080254.15383-1-tvrtko.ursulin@linux.intel.com --- drivers/gpu/drm/i915/Makefile | 3 +- drivers/gpu/drm/i915/selftests/igt_spinner.c | 199 ++++++++++++ drivers/gpu/drm/i915/selftests/igt_spinner.h | 37 +++ drivers/gpu/drm/i915/selftests/intel_lrc.c | 301 ++++--------------- 4 files changed, 290 insertions(+), 250 deletions(-) create mode 100644 drivers/gpu/drm/i915/selftests/igt_spinner.c create mode 100644 drivers/gpu/drm/i915/selftests/igt_spinner.h diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile index 8370b9de6e4f..50a8fa8fce64 100644 --- a/drivers/gpu/drm/i915/Makefile +++ b/drivers/gpu/drm/i915/Makefile @@ -165,7 +165,8 @@ i915-$(CONFIG_DRM_I915_CAPTURE_ERROR) += i915_gpu_error.o i915-$(CONFIG_DRM_I915_SELFTEST) += \ selftests/i915_random.o \ selftests/i915_selftest.o \ - selftests/igt_flush_test.o + selftests/igt_flush_test.o \ + selftests/igt_spinner.o # virtual gpu code i915-y += i915_vgpu.o diff --git a/drivers/gpu/drm/i915/selftests/igt_spinner.c b/drivers/gpu/drm/i915/selftests/igt_spinner.c new file mode 100644 index 000000000000..8cd34f6e6859 --- /dev/null +++ b/drivers/gpu/drm/i915/selftests/igt_spinner.c @@ -0,0 +1,199 @@ +/* + * SPDX-License-Identifier: MIT + * + * Copyright © 2018 Intel Corporation + */ + +#include "igt_spinner.h" + +int igt_spinner_init(struct igt_spinner *spin, struct drm_i915_private *i915) +{ + unsigned int mode; + void *vaddr; + int err; + + GEM_BUG_ON(INTEL_GEN(i915) < 8); + + memset(spin, 0, sizeof(*spin)); + spin->i915 = i915; + + spin->hws = i915_gem_object_create_internal(i915, PAGE_SIZE); + if (IS_ERR(spin->hws)) { + err = PTR_ERR(spin->hws); + goto err; + } + + spin->obj = i915_gem_object_create_internal(i915, PAGE_SIZE); + if (IS_ERR(spin->obj)) { + err = PTR_ERR(spin->obj); + goto err_hws; + } + + i915_gem_object_set_cache_level(spin->hws, I915_CACHE_LLC); + vaddr = i915_gem_object_pin_map(spin->hws, I915_MAP_WB); + if (IS_ERR(vaddr)) { + err = PTR_ERR(vaddr); + goto err_obj; + } + spin->seqno = memset(vaddr, 0xff, PAGE_SIZE); + + mode = i915_coherent_map_type(i915); + vaddr = i915_gem_object_pin_map(spin->obj, mode); + if (IS_ERR(vaddr)) { + err = PTR_ERR(vaddr); + goto err_unpin_hws; + } + spin->batch = vaddr; + + return 0; + +err_unpin_hws: + i915_gem_object_unpin_map(spin->hws); +err_obj: + i915_gem_object_put(spin->obj); +err_hws: + i915_gem_object_put(spin->hws); +err: + return err; +} + +static unsigned int seqno_offset(u64 fence) +{ + return offset_in_page(sizeof(u32) * fence); +} + +static u64 hws_address(const struct i915_vma *hws, + const struct i915_request *rq) +{ + return hws->node.start + seqno_offset(rq->fence.context); +} + +static int emit_recurse_batch(struct igt_spinner *spin, + struct i915_request *rq, + u32 arbitration_command) +{ + struct i915_address_space *vm = &rq->gem_context->ppgtt->vm; + struct i915_vma *hws, *vma; + u32 *batch; + int err; + + vma = i915_vma_instance(spin->obj, vm, NULL); + if (IS_ERR(vma)) + return PTR_ERR(vma); + + hws = i915_vma_instance(spin->hws, vm, NULL); + if (IS_ERR(hws)) + return PTR_ERR(hws); + + err = i915_vma_pin(vma, 0, 0, PIN_USER); + if (err) + return err; + + err = i915_vma_pin(hws, 0, 0, PIN_USER); + if (err) + goto unpin_vma; + + err = i915_vma_move_to_active(vma, rq, 0); + if (err) + goto unpin_hws; + + if (!i915_gem_object_has_active_reference(vma->obj)) { + i915_gem_object_get(vma->obj); + i915_gem_object_set_active_reference(vma->obj); + } + + err = i915_vma_move_to_active(hws, rq, 0); + if (err) + goto unpin_hws; + + if (!i915_gem_object_has_active_reference(hws->obj)) { + i915_gem_object_get(hws->obj); + i915_gem_object_set_active_reference(hws->obj); + } + + batch = spin->batch; + + *batch++ = MI_STORE_DWORD_IMM_GEN4; + *batch++ = lower_32_bits(hws_address(hws, rq)); + *batch++ = upper_32_bits(hws_address(hws, rq)); + *batch++ = rq->fence.seqno; + + *batch++ = arbitration_command; + + *batch++ = MI_BATCH_BUFFER_START | 1 << 8 | 1; + *batch++ = lower_32_bits(vma->node.start); + *batch++ = upper_32_bits(vma->node.start); + *batch++ = MI_BATCH_BUFFER_END; /* not reached */ + + i915_gem_chipset_flush(spin->i915); + + err = rq->engine->emit_bb_start(rq, vma->node.start, PAGE_SIZE, 0); + +unpin_hws: + i915_vma_unpin(hws); +unpin_vma: + i915_vma_unpin(vma); + return err; +} + +struct i915_request * +igt_spinner_create_request(struct igt_spinner *spin, + struct i915_gem_context *ctx, + struct intel_engine_cs *engine, + u32 arbitration_command) +{ + struct i915_request *rq; + int err; + + rq = i915_request_alloc(engine, ctx); + if (IS_ERR(rq)) + return rq; + + err = emit_recurse_batch(spin, rq, arbitration_command); + if (err) { + i915_request_add(rq); + return ERR_PTR(err); + } + + return rq; +} + +static u32 +hws_seqno(const struct igt_spinner *spin, const struct i915_request *rq) +{ + u32 *seqno = spin->seqno + seqno_offset(rq->fence.context); + + return READ_ONCE(*seqno); +} + +void igt_spinner_end(struct igt_spinner *spin) +{ + *spin->batch = MI_BATCH_BUFFER_END; + i915_gem_chipset_flush(spin->i915); +} + +void igt_spinner_fini(struct igt_spinner *spin) +{ + igt_spinner_end(spin); + + i915_gem_object_unpin_map(spin->obj); + i915_gem_object_put(spin->obj); + + i915_gem_object_unpin_map(spin->hws); + i915_gem_object_put(spin->hws); +} + +bool igt_wait_for_spinner(struct igt_spinner *spin, struct i915_request *rq) +{ + if (!wait_event_timeout(rq->execute, + READ_ONCE(rq->global_seqno), + msecs_to_jiffies(10))) + return false; + + return !(wait_for_us(i915_seqno_passed(hws_seqno(spin, rq), + rq->fence.seqno), + 10) && + wait_for(i915_seqno_passed(hws_seqno(spin, rq), + rq->fence.seqno), + 1000)); +} diff --git a/drivers/gpu/drm/i915/selftests/igt_spinner.h b/drivers/gpu/drm/i915/selftests/igt_spinner.h new file mode 100644 index 000000000000..391777c76dc7 --- /dev/null +++ b/drivers/gpu/drm/i915/selftests/igt_spinner.h @@ -0,0 +1,37 @@ +/* + * SPDX-License-Identifier: MIT + * + * Copyright © 2018 Intel Corporation + */ + +#ifndef __I915_SELFTESTS_IGT_SPINNER_H__ +#define __I915_SELFTESTS_IGT_SPINNER_H__ + +#include "../i915_selftest.h" + +#include "../i915_drv.h" +#include "../i915_request.h" +#include "../intel_ringbuffer.h" +#include "../i915_gem_context.h" + +struct igt_spinner { + struct drm_i915_private *i915; + struct drm_i915_gem_object *hws; + struct drm_i915_gem_object *obj; + u32 *batch; + void *seqno; +}; + +int igt_spinner_init(struct igt_spinner *spin, struct drm_i915_private *i915); +void igt_spinner_fini(struct igt_spinner *spin); + +struct i915_request * +igt_spinner_create_request(struct igt_spinner *spin, + struct i915_gem_context *ctx, + struct intel_engine_cs *engine, + u32 arbitration_command); +void igt_spinner_end(struct igt_spinner *spin); + +bool igt_wait_for_spinner(struct igt_spinner *spin, struct i915_request *rq); + +#endif diff --git a/drivers/gpu/drm/i915/selftests/intel_lrc.c b/drivers/gpu/drm/i915/selftests/intel_lrc.c index 94fc0e5c8766..ca461e3a5f27 100644 --- a/drivers/gpu/drm/i915/selftests/intel_lrc.c +++ b/drivers/gpu/drm/i915/selftests/intel_lrc.c @@ -6,216 +6,18 @@ #include "../i915_selftest.h" #include "igt_flush_test.h" +#include "igt_spinner.h" #include "i915_random.h" #include "mock_context.h" -struct spinner { - struct drm_i915_private *i915; - struct drm_i915_gem_object *hws; - struct drm_i915_gem_object *obj; - u32 *batch; - void *seqno; -}; - -static int spinner_init(struct spinner *spin, struct drm_i915_private *i915) -{ - unsigned int mode; - void *vaddr; - int err; - - GEM_BUG_ON(INTEL_GEN(i915) < 8); - - memset(spin, 0, sizeof(*spin)); - spin->i915 = i915; - - spin->hws = i915_gem_object_create_internal(i915, PAGE_SIZE); - if (IS_ERR(spin->hws)) { - err = PTR_ERR(spin->hws); - goto err; - } - - spin->obj = i915_gem_object_create_internal(i915, PAGE_SIZE); - if (IS_ERR(spin->obj)) { - err = PTR_ERR(spin->obj); - goto err_hws; - } - - i915_gem_object_set_cache_level(spin->hws, I915_CACHE_LLC); - vaddr = i915_gem_object_pin_map(spin->hws, I915_MAP_WB); - if (IS_ERR(vaddr)) { - err = PTR_ERR(vaddr); - goto err_obj; - } - spin->seqno = memset(vaddr, 0xff, PAGE_SIZE); - - mode = i915_coherent_map_type(i915); - vaddr = i915_gem_object_pin_map(spin->obj, mode); - if (IS_ERR(vaddr)) { - err = PTR_ERR(vaddr); - goto err_unpin_hws; - } - spin->batch = vaddr; - - return 0; - -err_unpin_hws: - i915_gem_object_unpin_map(spin->hws); -err_obj: - i915_gem_object_put(spin->obj); -err_hws: - i915_gem_object_put(spin->hws); -err: - return err; -} - -static unsigned int seqno_offset(u64 fence) -{ - return offset_in_page(sizeof(u32) * fence); -} - -static u64 hws_address(const struct i915_vma *hws, - const struct i915_request *rq) -{ - return hws->node.start + seqno_offset(rq->fence.context); -} - -static int emit_recurse_batch(struct spinner *spin, - struct i915_request *rq, - u32 arbitration_command) -{ - struct i915_address_space *vm = &rq->gem_context->ppgtt->vm; - struct i915_vma *hws, *vma; - u32 *batch; - int err; - - vma = i915_vma_instance(spin->obj, vm, NULL); - if (IS_ERR(vma)) - return PTR_ERR(vma); - - hws = i915_vma_instance(spin->hws, vm, NULL); - if (IS_ERR(hws)) - return PTR_ERR(hws); - - err = i915_vma_pin(vma, 0, 0, PIN_USER); - if (err) - return err; - - err = i915_vma_pin(hws, 0, 0, PIN_USER); - if (err) - goto unpin_vma; - - err = i915_vma_move_to_active(vma, rq, 0); - if (err) - goto unpin_hws; - - if (!i915_gem_object_has_active_reference(vma->obj)) { - i915_gem_object_get(vma->obj); - i915_gem_object_set_active_reference(vma->obj); - } - - err = i915_vma_move_to_active(hws, rq, 0); - if (err) - goto unpin_hws; - - if (!i915_gem_object_has_active_reference(hws->obj)) { - i915_gem_object_get(hws->obj); - i915_gem_object_set_active_reference(hws->obj); - } - - batch = spin->batch; - - *batch++ = MI_STORE_DWORD_IMM_GEN4; - *batch++ = lower_32_bits(hws_address(hws, rq)); - *batch++ = upper_32_bits(hws_address(hws, rq)); - *batch++ = rq->fence.seqno; - - *batch++ = arbitration_command; - - *batch++ = MI_BATCH_BUFFER_START | 1 << 8 | 1; - *batch++ = lower_32_bits(vma->node.start); - *batch++ = upper_32_bits(vma->node.start); - *batch++ = MI_BATCH_BUFFER_END; /* not reached */ - - i915_gem_chipset_flush(spin->i915); - - err = rq->engine->emit_bb_start(rq, vma->node.start, PAGE_SIZE, 0); - -unpin_hws: - i915_vma_unpin(hws); -unpin_vma: - i915_vma_unpin(vma); - return err; -} - -static struct i915_request * -spinner_create_request(struct spinner *spin, - struct i915_gem_context *ctx, - struct intel_engine_cs *engine, - u32 arbitration_command) -{ - struct i915_request *rq; - int err; - - rq = i915_request_alloc(engine, ctx); - if (IS_ERR(rq)) - return rq; - - err = emit_recurse_batch(spin, rq, arbitration_command); - if (err) { - i915_request_add(rq); - return ERR_PTR(err); - } - - return rq; -} - -static u32 hws_seqno(const struct spinner *spin, const struct i915_request *rq) -{ - u32 *seqno = spin->seqno + seqno_offset(rq->fence.context); - - return READ_ONCE(*seqno); -} - -static void spinner_end(struct spinner *spin) -{ - *spin->batch = MI_BATCH_BUFFER_END; - i915_gem_chipset_flush(spin->i915); -} - -static void spinner_fini(struct spinner *spin) -{ - spinner_end(spin); - - i915_gem_object_unpin_map(spin->obj); - i915_gem_object_put(spin->obj); - - i915_gem_object_unpin_map(spin->hws); - i915_gem_object_put(spin->hws); -} - -static bool wait_for_spinner(struct spinner *spin, struct i915_request *rq) -{ - if (!wait_event_timeout(rq->execute, - READ_ONCE(rq->global_seqno), - msecs_to_jiffies(10))) - return false; - - return !(wait_for_us(i915_seqno_passed(hws_seqno(spin, rq), - rq->fence.seqno), - 10) && - wait_for(i915_seqno_passed(hws_seqno(spin, rq), - rq->fence.seqno), - 1000)); -} - static int live_sanitycheck(void *arg) { struct drm_i915_private *i915 = arg; struct intel_engine_cs *engine; struct i915_gem_context *ctx; enum intel_engine_id id; - struct spinner spin; + struct igt_spinner spin; int err = -ENOMEM; if (!HAS_LOGICAL_RING_CONTEXTS(i915)) @@ -224,7 +26,7 @@ static int live_sanitycheck(void *arg) mutex_lock(&i915->drm.struct_mutex); intel_runtime_pm_get(i915); - if (spinner_init(&spin, i915)) + if (igt_spinner_init(&spin, i915)) goto err_unlock; ctx = kernel_context(i915); @@ -234,14 +36,14 @@ static int live_sanitycheck(void *arg) for_each_engine(engine, i915, id) { struct i915_request *rq; - rq = spinner_create_request(&spin, ctx, engine, MI_NOOP); + rq = igt_spinner_create_request(&spin, ctx, engine, MI_NOOP); if (IS_ERR(rq)) { err = PTR_ERR(rq); goto err_ctx; } i915_request_add(rq); - if (!wait_for_spinner(&spin, rq)) { + if (!igt_wait_for_spinner(&spin, rq)) { GEM_TRACE("spinner failed to start\n"); GEM_TRACE_DUMP(); i915_gem_set_wedged(i915); @@ -249,7 +51,7 @@ static int live_sanitycheck(void *arg) goto err_ctx; } - spinner_end(&spin); + igt_spinner_end(&spin); if (igt_flush_test(i915, I915_WAIT_LOCKED)) { err = -EIO; goto err_ctx; @@ -260,7 +62,7 @@ static int live_sanitycheck(void *arg) err_ctx: kernel_context_close(ctx); err_spin: - spinner_fini(&spin); + igt_spinner_fini(&spin); err_unlock: igt_flush_test(i915, I915_WAIT_LOCKED); intel_runtime_pm_put(i915); @@ -272,7 +74,7 @@ static int live_preempt(void *arg) { struct drm_i915_private *i915 = arg; struct i915_gem_context *ctx_hi, *ctx_lo; - struct spinner spin_hi, spin_lo; + struct igt_spinner spin_hi, spin_lo; struct intel_engine_cs *engine; enum intel_engine_id id; int err = -ENOMEM; @@ -283,10 +85,10 @@ static int live_preempt(void *arg) mutex_lock(&i915->drm.struct_mutex); intel_runtime_pm_get(i915); - if (spinner_init(&spin_hi, i915)) + if (igt_spinner_init(&spin_hi, i915)) goto err_unlock; - if (spinner_init(&spin_lo, i915)) + if (igt_spinner_init(&spin_lo, i915)) goto err_spin_hi; ctx_hi = kernel_context(i915); @@ -304,15 +106,15 @@ static int live_preempt(void *arg) for_each_engine(engine, i915, id) { struct i915_request *rq; - rq = spinner_create_request(&spin_lo, ctx_lo, engine, - MI_ARB_CHECK); + rq = igt_spinner_create_request(&spin_lo, ctx_lo, engine, + MI_ARB_CHECK); if (IS_ERR(rq)) { err = PTR_ERR(rq); goto err_ctx_lo; } i915_request_add(rq); - if (!wait_for_spinner(&spin_lo, rq)) { + if (!igt_wait_for_spinner(&spin_lo, rq)) { GEM_TRACE("lo spinner failed to start\n"); GEM_TRACE_DUMP(); i915_gem_set_wedged(i915); @@ -320,16 +122,16 @@ static int live_preempt(void *arg) goto err_ctx_lo; } - rq = spinner_create_request(&spin_hi, ctx_hi, engine, - MI_ARB_CHECK); + rq = igt_spinner_create_request(&spin_hi, ctx_hi, engine, + MI_ARB_CHECK); if (IS_ERR(rq)) { - spinner_end(&spin_lo); + igt_spinner_end(&spin_lo); err = PTR_ERR(rq); goto err_ctx_lo; } i915_request_add(rq); - if (!wait_for_spinner(&spin_hi, rq)) { + if (!igt_wait_for_spinner(&spin_hi, rq)) { GEM_TRACE("hi spinner failed to start\n"); GEM_TRACE_DUMP(); i915_gem_set_wedged(i915); @@ -337,8 +139,8 @@ static int live_preempt(void *arg) goto err_ctx_lo; } - spinner_end(&spin_hi); - spinner_end(&spin_lo); + igt_spinner_end(&spin_hi); + igt_spinner_end(&spin_lo); if (igt_flush_test(i915, I915_WAIT_LOCKED)) { err = -EIO; goto err_ctx_lo; @@ -351,9 +153,9 @@ err_ctx_lo: err_ctx_hi: kernel_context_close(ctx_hi); err_spin_lo: - spinner_fini(&spin_lo); + igt_spinner_fini(&spin_lo); err_spin_hi: - spinner_fini(&spin_hi); + igt_spinner_fini(&spin_hi); err_unlock: igt_flush_test(i915, I915_WAIT_LOCKED); intel_runtime_pm_put(i915); @@ -365,7 +167,7 @@ static int live_late_preempt(void *arg) { struct drm_i915_private *i915 = arg; struct i915_gem_context *ctx_hi, *ctx_lo; - struct spinner spin_hi, spin_lo; + struct igt_spinner spin_hi, spin_lo; struct intel_engine_cs *engine; struct i915_sched_attr attr = {}; enum intel_engine_id id; @@ -377,10 +179,10 @@ static int live_late_preempt(void *arg) mutex_lock(&i915->drm.struct_mutex); intel_runtime_pm_get(i915); - if (spinner_init(&spin_hi, i915)) + if (igt_spinner_init(&spin_hi, i915)) goto err_unlock; - if (spinner_init(&spin_lo, i915)) + if (igt_spinner_init(&spin_lo, i915)) goto err_spin_hi; ctx_hi = kernel_context(i915); @@ -394,28 +196,29 @@ static int live_late_preempt(void *arg) for_each_engine(engine, i915, id) { struct i915_request *rq; - rq = spinner_create_request(&spin_lo, ctx_lo, engine, - MI_ARB_CHECK); + rq = igt_spinner_create_request(&spin_lo, ctx_lo, engine, + MI_ARB_CHECK); if (IS_ERR(rq)) { err = PTR_ERR(rq); goto err_ctx_lo; } i915_request_add(rq); - if (!wait_for_spinner(&spin_lo, rq)) { + if (!igt_wait_for_spinner(&spin_lo, rq)) { pr_err("First context failed to start\n"); goto err_wedged; } - rq = spinner_create_request(&spin_hi, ctx_hi, engine, MI_NOOP); + rq = igt_spinner_create_request(&spin_hi, ctx_hi, engine, + MI_NOOP); if (IS_ERR(rq)) { - spinner_end(&spin_lo); + igt_spinner_end(&spin_lo); err = PTR_ERR(rq); goto err_ctx_lo; } i915_request_add(rq); - if (wait_for_spinner(&spin_hi, rq)) { + if (igt_wait_for_spinner(&spin_hi, rq)) { pr_err("Second context overtook first?\n"); goto err_wedged; } @@ -423,14 +226,14 @@ static int live_late_preempt(void *arg) attr.priority = I915_USER_PRIORITY(I915_PRIORITY_MAX); engine->schedule(rq, &attr); - if (!wait_for_spinner(&spin_hi, rq)) { + if (!igt_wait_for_spinner(&spin_hi, rq)) { pr_err("High priority context failed to preempt the low priority context\n"); GEM_TRACE_DUMP(); goto err_wedged; } - spinner_end(&spin_hi); - spinner_end(&spin_lo); + igt_spinner_end(&spin_hi); + igt_spinner_end(&spin_lo); if (igt_flush_test(i915, I915_WAIT_LOCKED)) { err = -EIO; goto err_ctx_lo; @@ -443,9 +246,9 @@ err_ctx_lo: err_ctx_hi: kernel_context_close(ctx_hi); err_spin_lo: - spinner_fini(&spin_lo); + igt_spinner_fini(&spin_lo); err_spin_hi: - spinner_fini(&spin_hi); + igt_spinner_fini(&spin_hi); err_unlock: igt_flush_test(i915, I915_WAIT_LOCKED); intel_runtime_pm_put(i915); @@ -453,8 +256,8 @@ err_unlock: return err; err_wedged: - spinner_end(&spin_hi); - spinner_end(&spin_lo); + igt_spinner_end(&spin_hi); + igt_spinner_end(&spin_lo); i915_gem_set_wedged(i915); err = -EIO; goto err_ctx_lo; @@ -464,7 +267,7 @@ static int live_preempt_hang(void *arg) { struct drm_i915_private *i915 = arg; struct i915_gem_context *ctx_hi, *ctx_lo; - struct spinner spin_hi, spin_lo; + struct igt_spinner spin_hi, spin_lo; struct intel_engine_cs *engine; enum intel_engine_id id; int err = -ENOMEM; @@ -478,10 +281,10 @@ static int live_preempt_hang(void *arg) mutex_lock(&i915->drm.struct_mutex); intel_runtime_pm_get(i915); - if (spinner_init(&spin_hi, i915)) + if (igt_spinner_init(&spin_hi, i915)) goto err_unlock; - if (spinner_init(&spin_lo, i915)) + if (igt_spinner_init(&spin_lo, i915)) goto err_spin_hi; ctx_hi = kernel_context(i915); @@ -500,15 +303,15 @@ static int live_preempt_hang(void *arg) if (!intel_engine_has_preemption(engine)) continue; - rq = spinner_create_request(&spin_lo, ctx_lo, engine, - MI_ARB_CHECK); + rq = igt_spinner_create_request(&spin_lo, ctx_lo, engine, + MI_ARB_CHECK); if (IS_ERR(rq)) { err = PTR_ERR(rq); goto err_ctx_lo; } i915_request_add(rq); - if (!wait_for_spinner(&spin_lo, rq)) { + if (!igt_wait_for_spinner(&spin_lo, rq)) { GEM_TRACE("lo spinner failed to start\n"); GEM_TRACE_DUMP(); i915_gem_set_wedged(i915); @@ -516,10 +319,10 @@ static int live_preempt_hang(void *arg) goto err_ctx_lo; } - rq = spinner_create_request(&spin_hi, ctx_hi, engine, - MI_ARB_CHECK); + rq = igt_spinner_create_request(&spin_hi, ctx_hi, engine, + MI_ARB_CHECK); if (IS_ERR(rq)) { - spinner_end(&spin_lo); + igt_spinner_end(&spin_lo); err = PTR_ERR(rq); goto err_ctx_lo; } @@ -544,7 +347,7 @@ static int live_preempt_hang(void *arg) engine->execlists.preempt_hang.inject_hang = false; - if (!wait_for_spinner(&spin_hi, rq)) { + if (!igt_wait_for_spinner(&spin_hi, rq)) { GEM_TRACE("hi spinner failed to start\n"); GEM_TRACE_DUMP(); i915_gem_set_wedged(i915); @@ -552,8 +355,8 @@ static int live_preempt_hang(void *arg) goto err_ctx_lo; } - spinner_end(&spin_hi); - spinner_end(&spin_lo); + igt_spinner_end(&spin_hi); + igt_spinner_end(&spin_lo); if (igt_flush_test(i915, I915_WAIT_LOCKED)) { err = -EIO; goto err_ctx_lo; @@ -566,9 +369,9 @@ err_ctx_lo: err_ctx_hi: kernel_context_close(ctx_hi); err_spin_lo: - spinner_fini(&spin_lo); + igt_spinner_fini(&spin_lo); err_spin_hi: - spinner_fini(&spin_hi); + igt_spinner_fini(&spin_hi); err_unlock: igt_flush_test(i915, I915_WAIT_LOCKED); intel_runtime_pm_put(i915); From b9f78d675230a9f65611600496235cdbd369d55d Mon Sep 17 00:00:00 2001 From: Tvrtko Ursulin Date: Fri, 30 Nov 2018 09:52:11 +0000 Subject: [PATCH 55/98] drm/i915/selftests: Fix live_workarounds to actually do resets The test was missing some magic ingredients to actually trigger the resets. In case of the full reset we need the I915_RESET_HANDOFF flag set, and in case of engine reset we need a busy request. Thanks to Chris for helping with reset magic. v2: * Grab RPM ref over reset. Signed-off-by: Tvrtko Ursulin Cc: Chris Wilson Reviewed-by: Chris Wilson Link: https://patchwork.freedesktop.org/patch/msgid/20181130095211.23849-1-tvrtko.ursulin@linux.intel.com --- .../drm/i915/selftests/intel_workarounds.c | 65 +++++++++++++++---- 1 file changed, 53 insertions(+), 12 deletions(-) diff --git a/drivers/gpu/drm/i915/selftests/intel_workarounds.c b/drivers/gpu/drm/i915/selftests/intel_workarounds.c index d1a0923d2f38..80396b3592f5 100644 --- a/drivers/gpu/drm/i915/selftests/intel_workarounds.c +++ b/drivers/gpu/drm/i915/selftests/intel_workarounds.c @@ -6,6 +6,7 @@ #include "../i915_selftest.h" +#include "igt_spinner.h" #include "igt_wedge_me.h" #include "mock_context.h" @@ -159,35 +160,57 @@ out_put: static int do_device_reset(struct intel_engine_cs *engine) { - i915_reset(engine->i915, ENGINE_MASK(engine->id), NULL); + set_bit(I915_RESET_HANDOFF, &engine->i915->gpu_error.flags); + i915_reset(engine->i915, ENGINE_MASK(engine->id), "live_workarounds"); return 0; } static int do_engine_reset(struct intel_engine_cs *engine) { - return i915_reset_engine(engine, NULL); + return i915_reset_engine(engine, "live_workarounds"); } -static int switch_to_scratch_context(struct intel_engine_cs *engine) +static int +switch_to_scratch_context(struct intel_engine_cs *engine, + struct igt_spinner *spin) { struct i915_gem_context *ctx; struct i915_request *rq; + int err = 0; ctx = kernel_context(engine->i915); if (IS_ERR(ctx)) return PTR_ERR(ctx); intel_runtime_pm_get(engine->i915); - rq = i915_request_alloc(engine, ctx); + + if (spin) + rq = igt_spinner_create_request(spin, ctx, engine, MI_NOOP); + else + rq = i915_request_alloc(engine, ctx); + intel_runtime_pm_put(engine->i915); kernel_context_close(ctx); - if (IS_ERR(rq)) - return PTR_ERR(rq); + + if (IS_ERR(rq)) { + spin = NULL; + err = PTR_ERR(rq); + goto err; + } i915_request_add(rq); - return 0; + if (spin && !igt_wait_for_spinner(spin, rq)) { + pr_err("Spinner failed to start\n"); + err = -ETIMEDOUT; + } + +err: + if (err && spin) + igt_spinner_end(spin); + + return err; } static int check_whitelist_across_reset(struct intel_engine_cs *engine, @@ -195,10 +218,22 @@ static int check_whitelist_across_reset(struct intel_engine_cs *engine, const struct whitelist *w, const char *name) { + struct drm_i915_private *i915 = engine->i915; + bool want_spin = reset == do_engine_reset; struct i915_gem_context *ctx; + struct igt_spinner spin; int err; - ctx = kernel_context(engine->i915); + pr_info("Checking %d whitelisted registers (RING_NONPRIV) [%s]\n", + w->count, name); + + if (want_spin) { + err = igt_spinner_init(&spin, i915); + if (err) + return err; + } + + ctx = kernel_context(i915); if (IS_ERR(ctx)) return PTR_ERR(ctx); @@ -208,11 +243,19 @@ static int check_whitelist_across_reset(struct intel_engine_cs *engine, goto out; } - err = switch_to_scratch_context(engine); + err = switch_to_scratch_context(engine, want_spin ? &spin : NULL); if (err) goto out; + intel_runtime_pm_get(i915); err = reset(engine); + intel_runtime_pm_put(i915); + + if (want_spin) { + igt_spinner_end(&spin); + igt_spinner_fini(&spin); + } + if (err) { pr_err("%s reset failed\n", name); goto out; @@ -227,7 +270,7 @@ static int check_whitelist_across_reset(struct intel_engine_cs *engine, kernel_context_close(ctx); - ctx = kernel_context(engine->i915); + ctx = kernel_context(i915); if (IS_ERR(ctx)) return PTR_ERR(ctx); @@ -259,8 +302,6 @@ static int live_reset_whitelist(void *arg) if (!whitelist_build(engine, &w)) return 0; - pr_info("Checking %d whitelisted registers (RING_NONPRIV)\n", w.count); - set_bit(I915_RESET_BACKOFF, &error->flags); set_bit(I915_RESET_ENGINE + engine->id, &error->flags); From f0ad62a631e040ae4413286a4b46a90c5ce42d07 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Jos=C3=A9=20Roberto=20de=20Souza?= Date: Tue, 27 Nov 2018 23:28:38 -0800 Subject: [PATCH 56/98] drm/i915/psr: Get pipe id following atomic guidelines MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit As stated in struct drm_encoder, crtc field should only be used by non-atomic drivers. So here caching the pipe id in intel_psr_enable() what is way more simple and efficient than at every call to intel_psr_flush()/invalidate() get the drm.mode_config.connection_mutex lock to safely be able to get the pipe id by reading drm_connector_state.crtc. This should fix the null pointer dereference crash below as the previous way to get the pipe id was prone to race conditions. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105959 Cc: Dhinakaran Pandiyan Cc: Rodrigo Vivi Signed-off-by: José Roberto de Souza Reviewed-by: Rodrigo Vivi Link: https://patchwork.freedesktop.org/patch/msgid/20181128072838.22773-1-jose.souza@intel.com --- drivers/gpu/drm/i915/i915_drv.h | 1 + drivers/gpu/drm/i915/intel_psr.c | 19 ++++--------------- 2 files changed, 5 insertions(+), 15 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 43ac6873a2bb..d45475287130 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -495,6 +495,7 @@ struct i915_psr { bool sink_support; bool prepared, enabled; struct intel_dp *dp; + enum pipe pipe; bool active; struct work_struct work; unsigned busy_frontbuffer_bits; diff --git a/drivers/gpu/drm/i915/intel_psr.c b/drivers/gpu/drm/i915/intel_psr.c index 2084784f320d..419e56342523 100644 --- a/drivers/gpu/drm/i915/intel_psr.c +++ b/drivers/gpu/drm/i915/intel_psr.c @@ -719,6 +719,7 @@ void intel_psr_enable(struct intel_dp *intel_dp, dev_priv->psr.psr2_enabled = intel_psr2_enabled(dev_priv, crtc_state); dev_priv->psr.busy_frontbuffer_bits = 0; dev_priv->psr.prepared = true; + dev_priv->psr.pipe = to_intel_crtc(crtc_state->base.crtc)->pipe; if (psr_global_enabled(dev_priv->psr.debug)) intel_psr_enable_locked(dev_priv, crtc_state); @@ -1026,9 +1027,6 @@ unlock: void intel_psr_invalidate(struct drm_i915_private *dev_priv, unsigned frontbuffer_bits, enum fb_op_origin origin) { - struct drm_crtc *crtc; - enum pipe pipe; - if (!CAN_PSR(dev_priv)) return; @@ -1041,10 +1039,7 @@ void intel_psr_invalidate(struct drm_i915_private *dev_priv, return; } - crtc = dp_to_dig_port(dev_priv->psr.dp)->base.base.crtc; - pipe = to_intel_crtc(crtc)->pipe; - - frontbuffer_bits &= INTEL_FRONTBUFFER_ALL_MASK(pipe); + frontbuffer_bits &= INTEL_FRONTBUFFER_ALL_MASK(dev_priv->psr.pipe); dev_priv->psr.busy_frontbuffer_bits |= frontbuffer_bits; if (frontbuffer_bits) @@ -1069,9 +1064,6 @@ void intel_psr_invalidate(struct drm_i915_private *dev_priv, void intel_psr_flush(struct drm_i915_private *dev_priv, unsigned frontbuffer_bits, enum fb_op_origin origin) { - struct drm_crtc *crtc; - enum pipe pipe; - if (!CAN_PSR(dev_priv)) return; @@ -1084,10 +1076,7 @@ void intel_psr_flush(struct drm_i915_private *dev_priv, return; } - crtc = dp_to_dig_port(dev_priv->psr.dp)->base.base.crtc; - pipe = to_intel_crtc(crtc)->pipe; - - frontbuffer_bits &= INTEL_FRONTBUFFER_ALL_MASK(pipe); + frontbuffer_bits &= INTEL_FRONTBUFFER_ALL_MASK(dev_priv->psr.pipe); dev_priv->psr.busy_frontbuffer_bits &= ~frontbuffer_bits; /* By definition flush = invalidate + flush */ @@ -1101,7 +1090,7 @@ void intel_psr_flush(struct drm_i915_private *dev_priv, * but it makes more sense write to the current active * pipe. */ - I915_WRITE(CURSURFLIVE(pipe), 0); + I915_WRITE(CURSURFLIVE(dev_priv->psr.pipe), 0); } if (!dev_priv->psr.active && !dev_priv->psr.busy_frontbuffer_bits) From f36c071f6344e0a335ed4b4e0b3a38c0dd54648b Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Mon, 3 Dec 2018 11:36:56 +0000 Subject: [PATCH 57/98] drm/i915/ringbuffer: Clear semaphore sync registers on ring init Ensure that the sync registers are cleared every time we restart the ring to avoid stale values from creeping in from random neutrinos. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=108888 Signed-off-by: Chris Wilson Reviewed-by: Mika Kuoppala Link: https://patchwork.freedesktop.org/patch/msgid/20181203113701.12106-3-chris@chris-wilson.co.uk --- drivers/gpu/drm/i915/intel_ringbuffer.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index 28ae1e436ea6..d81eaf5f6b3e 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -529,6 +529,13 @@ static int init_ring_common(struct intel_engine_cs *engine) intel_engine_reset_breadcrumbs(engine); + if (HAS_LEGACY_SEMAPHORES(engine->i915)) { + I915_WRITE(RING_SYNC_0(engine->mmio_base), 0); + I915_WRITE(RING_SYNC_1(engine->mmio_base), 0); + if (HAS_VEBOX(dev_priv)) + I915_WRITE(RING_SYNC_2(engine->mmio_base), 0); + } + /* Enforce ordering by reading HEAD register back */ I915_READ_HEAD(engine); From f81ff31cb47162a748fc4cccb7bf3a01f2b6f2f0 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Mon, 3 Dec 2018 11:36:59 +0000 Subject: [PATCH 58/98] drm/i915/selftests: Terminate hangcheck sanitycheck forcibly If all else fails and we are stuck eternally waiting for the undying request, abandon all hope. Signed-off-by: Chris Wilson Reviewed-by: Mika Kuoppala Link: https://patchwork.freedesktop.org/patch/msgid/20181203113701.12106-6-chris@chris-wilson.co.uk --- drivers/gpu/drm/i915/selftests/intel_hangcheck.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/i915/selftests/intel_hangcheck.c b/drivers/gpu/drm/i915/selftests/intel_hangcheck.c index defe671130ab..a48fbe2557ea 100644 --- a/drivers/gpu/drm/i915/selftests/intel_hangcheck.c +++ b/drivers/gpu/drm/i915/selftests/intel_hangcheck.c @@ -308,6 +308,7 @@ static int igt_hang_sanitycheck(void *arg) goto unlock; for_each_engine(engine, i915, id) { + struct igt_wedge_me w; long timeout; if (!intel_engine_can_store_dword(engine)) @@ -328,9 +329,14 @@ static int igt_hang_sanitycheck(void *arg) i915_request_add(rq); - timeout = i915_request_wait(rq, - I915_WAIT_LOCKED, - MAX_SCHEDULE_TIMEOUT); + timeout = 0; + igt_wedge_on_timeout(&w, i915, HZ / 10 /* 100ms timeout*/) + timeout = i915_request_wait(rq, + I915_WAIT_LOCKED, + MAX_SCHEDULE_TIMEOUT); + if (i915_terminally_wedged(&i915->gpu_error)) + timeout = -EIO; + i915_request_put(rq); if (timeout < 0) { From 3b8c0d5bc9f48e3f2c2ecde08117f93fb6937d1a Mon Sep 17 00:00:00 2001 From: Jani Nikula Date: Thu, 29 Nov 2018 16:12:16 +0200 Subject: [PATCH 59/98] drm/i915/icl: push pll to port mapping/unmapping to ddi encoder hooks MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Unclutter the haswell_crtc_enable() and haswell_crtc_disable() functions a bit by moving the pll to port mapping and unmapping functions to the ddi encoder hooks. This allows removal of a bunch of boilerplate code from the functions. Additionally, the ICL DSI encoder needs to do the clock gating and ungating slightly differently, and this allows its own handling in a clean fashion. Cc: Madhav Chauhan Cc: Paulo Zanoni Cc: Ville Syrjälä Reviewed-by: Madhav Chauhan Reviewed-by: Ville Syrjälä Signed-off-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/f8e2982ceea4c05dc254a0c15e2b3be1d5f271d3.1543500285.git.jani.nikula@intel.com --- drivers/gpu/drm/i915/intel_ddi.c | 84 +++++++++++----------------- drivers/gpu/drm/i915/intel_display.c | 6 -- drivers/gpu/drm/i915/intel_drv.h | 6 -- 3 files changed, 34 insertions(+), 62 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c index 61d7145f93bf..d37ac95bc23d 100644 --- a/drivers/gpu/drm/i915/intel_ddi.c +++ b/drivers/gpu/drm/i915/intel_ddi.c @@ -2791,69 +2791,45 @@ uint32_t icl_dpclka_cfgcr0_clk_off(struct drm_i915_private *dev_priv, return 0; } -void icl_map_plls_to_ports(struct drm_crtc *crtc, - struct intel_crtc_state *crtc_state, - struct drm_atomic_state *old_state) +static void icl_map_plls_to_ports(struct intel_encoder *encoder, + const struct intel_crtc_state *crtc_state) { + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); struct intel_shared_dpll *pll = crtc_state->shared_dpll; - struct drm_i915_private *dev_priv = to_i915(crtc->dev); - struct drm_connector_state *conn_state; - struct drm_connector *conn; - int i; + enum port port = encoder->port; + u32 val; - for_each_new_connector_in_state(old_state, conn, conn_state, i) { - struct intel_encoder *encoder = - to_intel_encoder(conn_state->best_encoder); - enum port port; - uint32_t val; + mutex_lock(&dev_priv->dpll_lock); - if (conn_state->crtc != crtc) - continue; + val = I915_READ(DPCLKA_CFGCR0_ICL); + WARN_ON((val & icl_dpclka_cfgcr0_clk_off(dev_priv, port)) == 0); - port = encoder->port; - mutex_lock(&dev_priv->dpll_lock); - - val = I915_READ(DPCLKA_CFGCR0_ICL); - WARN_ON((val & icl_dpclka_cfgcr0_clk_off(dev_priv, port)) == 0); - - if (intel_port_is_combophy(dev_priv, port)) { - val &= ~DPCLKA_CFGCR0_DDI_CLK_SEL_MASK(port); - val |= DPCLKA_CFGCR0_DDI_CLK_SEL(pll->info->id, port); - I915_WRITE(DPCLKA_CFGCR0_ICL, val); - POSTING_READ(DPCLKA_CFGCR0_ICL); - } - - val &= ~icl_dpclka_cfgcr0_clk_off(dev_priv, port); + if (intel_port_is_combophy(dev_priv, port)) { + val &= ~DPCLKA_CFGCR0_DDI_CLK_SEL_MASK(port); + val |= DPCLKA_CFGCR0_DDI_CLK_SEL(pll->info->id, port); I915_WRITE(DPCLKA_CFGCR0_ICL, val); - - mutex_unlock(&dev_priv->dpll_lock); + POSTING_READ(DPCLKA_CFGCR0_ICL); } + + val &= ~icl_dpclka_cfgcr0_clk_off(dev_priv, port); + I915_WRITE(DPCLKA_CFGCR0_ICL, val); + + mutex_unlock(&dev_priv->dpll_lock); } -void icl_unmap_plls_to_ports(struct drm_crtc *crtc, - struct intel_crtc_state *crtc_state, - struct drm_atomic_state *old_state) +static void icl_unmap_plls_to_ports(struct intel_encoder *encoder) { - struct drm_i915_private *dev_priv = to_i915(crtc->dev); - struct drm_connector_state *old_conn_state; - struct drm_connector *conn; - int i; + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); + enum port port = encoder->port; + u32 val; - for_each_old_connector_in_state(old_state, conn, old_conn_state, i) { - struct intel_encoder *encoder = - to_intel_encoder(old_conn_state->best_encoder); - enum port port; + mutex_lock(&dev_priv->dpll_lock); - if (old_conn_state->crtc != crtc) - continue; + val = I915_READ(DPCLKA_CFGCR0_ICL); + val |= icl_dpclka_cfgcr0_clk_off(dev_priv, port); + I915_WRITE(DPCLKA_CFGCR0_ICL, val); - port = encoder->port; - mutex_lock(&dev_priv->dpll_lock); - I915_WRITE(DPCLKA_CFGCR0_ICL, - I915_READ(DPCLKA_CFGCR0_ICL) | - icl_dpclka_cfgcr0_clk_off(dev_priv, port)); - mutex_unlock(&dev_priv->dpll_lock); - } + mutex_unlock(&dev_priv->dpll_lock); } void icl_sanitize_encoder_pll_mapping(struct intel_encoder *encoder) @@ -3268,6 +3244,9 @@ static void intel_ddi_pre_enable(struct intel_encoder *encoder, WARN_ON(crtc_state->has_pch_encoder); + if (INTEL_GEN(dev_priv) >= 11) + icl_map_plls_to_ports(encoder, crtc_state); + intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, true); if (intel_crtc_has_type(crtc_state, INTEL_OUTPUT_HDMI)) { @@ -3370,6 +3349,8 @@ static void intel_ddi_post_disable(struct intel_encoder *encoder, const struct intel_crtc_state *old_crtc_state, const struct drm_connector_state *old_conn_state) { + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); + /* * When called from DP MST code: * - old_conn_state will be NULL @@ -3389,6 +3370,9 @@ static void intel_ddi_post_disable(struct intel_encoder *encoder, else intel_ddi_post_disable_dp(encoder, old_crtc_state, old_conn_state); + + if (INTEL_GEN(dev_priv) >= 11) + icl_unmap_plls_to_ports(encoder); } void intel_ddi_fdi_post_disable(struct intel_encoder *encoder, diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 789f647bd598..e349e2661a12 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -5726,9 +5726,6 @@ static void haswell_crtc_enable(struct intel_crtc_state *pipe_config, if (pipe_config->shared_dpll) intel_enable_shared_dpll(pipe_config); - if (INTEL_GEN(dev_priv) >= 11) - icl_map_plls_to_ports(crtc, pipe_config, old_state); - intel_encoders_pre_enable(crtc, pipe_config, old_state); if (intel_crtc_has_dp_encoder(pipe_config)) @@ -5932,9 +5929,6 @@ static void haswell_crtc_disable(struct intel_crtc_state *old_crtc_state, intel_encoders_post_disable(crtc, old_crtc_state, old_state); - if (INTEL_GEN(dev_priv) >= 11) - icl_unmap_plls_to_ports(crtc, old_crtc_state, old_state); - intel_encoders_post_pll_disable(crtc, old_crtc_state, old_state); } diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index 86ff57738cd9..04b9545994f2 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -1534,12 +1534,6 @@ u8 intel_ddi_dp_pre_emphasis_max(struct intel_encoder *encoder, u8 voltage_swing); int intel_ddi_toggle_hdcp_signalling(struct intel_encoder *intel_encoder, bool enable); -void icl_map_plls_to_ports(struct drm_crtc *crtc, - struct intel_crtc_state *crtc_state, - struct drm_atomic_state *old_state); -void icl_unmap_plls_to_ports(struct drm_crtc *crtc, - struct intel_crtc_state *crtc_state, - struct drm_atomic_state *old_state); void icl_sanitize_encoder_pll_mapping(struct intel_encoder *encoder); unsigned int intel_fb_align_height(const struct drm_framebuffer *fb, From 1dd07e56a3f1b38a68f3dbd263a4badc53ae274a Mon Sep 17 00:00:00 2001 From: Imre Deak Date: Thu, 29 Nov 2018 16:12:17 +0200 Subject: [PATCH 60/98] drm/i915/icl: Sanitize DDI port clock gating for DSI ports MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The requirement for the DDI port clock gating for a port in DSI mode is the opposite wrt. the case when the port is in DDI mode: the clock should be gated when the port is active and ungated when the port is inactive. Note that we cannot simply keep the DDI clock gated when the port will be only used in DSI mode: it must be gated/ungated at a specific spot in the DSI enable/disable sequence. Ensure the above for all ports of a DSI encoder, also adding a sanity check that we haven't registered another encoder using the same port (VBT should never allow this to happen). Cc: Madhav Chauhan Cc: Vandita Kulkarni Cc: Jani Nikula Cc: Ville Syrjälä Cc: Clint Taylor Reviewed-by: Jani Nikula Signed-off-by: Imre Deak Signed-off-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/ceb14d5a68e8e23988d923d4290a4f981789e616.1543500285.git.jani.nikula@intel.com --- drivers/gpu/drm/i915/intel_ddi.c | 65 +++++++++++++++++++++++--------- drivers/gpu/drm/i915/intel_dsi.h | 5 +++ 2 files changed, 53 insertions(+), 17 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c index d37ac95bc23d..6308d0f151e1 100644 --- a/drivers/gpu/drm/i915/intel_ddi.c +++ b/drivers/gpu/drm/i915/intel_ddi.c @@ -28,6 +28,7 @@ #include #include "i915_drv.h" #include "intel_drv.h" +#include "intel_dsi.h" struct ddi_buf_trans { u32 trans1; /* balance leg enable, de-emph level */ @@ -2836,8 +2837,9 @@ void icl_sanitize_encoder_pll_mapping(struct intel_encoder *encoder) { struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); u32 val; - enum port port = encoder->port; - bool clk_enabled; + enum port port; + u32 port_mask; + bool ddi_clk_needed; /* * In case of DP MST, we sanitize the primary encoder only, not the @@ -2846,9 +2848,6 @@ void icl_sanitize_encoder_pll_mapping(struct intel_encoder *encoder) if (encoder->type == INTEL_OUTPUT_DP_MST) return; - val = I915_READ(DPCLKA_CFGCR0_ICL); - clk_enabled = !(val & icl_dpclka_cfgcr0_clk_off(dev_priv, port)); - if (!encoder->base.crtc && intel_encoder_is_dp(encoder)) { u8 pipe_mask; bool is_mst; @@ -2862,20 +2861,52 @@ void icl_sanitize_encoder_pll_mapping(struct intel_encoder *encoder) return; } - if (clk_enabled == !!encoder->base.crtc) - return; + port_mask = BIT(encoder->port); + ddi_clk_needed = encoder->base.crtc; - /* - * Punt on the case now where clock is disabled, but the encoder is - * enabled, something else is really broken then. - */ - if (WARN_ON(!clk_enabled)) - return; + if (encoder->type == INTEL_OUTPUT_DSI) { + struct intel_encoder *other_encoder; - DRM_NOTE("Port %c is disabled but it has a mapped PLL, unmap it\n", - port_name(port)); - val |= icl_dpclka_cfgcr0_clk_off(dev_priv, port); - I915_WRITE(DPCLKA_CFGCR0_ICL, val); + port_mask = intel_dsi_encoder_ports(encoder); + /* + * Sanity check that we haven't incorrectly registered another + * encoder using any of the ports of this DSI encoder. + */ + for_each_intel_encoder(&dev_priv->drm, other_encoder) { + if (other_encoder == encoder) + continue; + + if (WARN_ON(port_mask & BIT(other_encoder->port))) + return; + } + /* + * DSI ports should have their DDI clock ungated when disabled + * and gated when enabled. + */ + ddi_clk_needed = !encoder->base.crtc; + } + + val = I915_READ(DPCLKA_CFGCR0_ICL); + for_each_port_masked(port, port_mask) { + bool ddi_clk_ungated = !(val & + icl_dpclka_cfgcr0_clk_off(dev_priv, + port)); + + if (ddi_clk_needed == ddi_clk_ungated) + continue; + + /* + * Punt on the case now where clock is gated, but it would + * be needed by the port. Something else is really broken then. + */ + if (WARN_ON(ddi_clk_needed)) + continue; + + DRM_NOTE("Port %c is disabled/in DSI mode with an ungated DDI clock, gate it\n", + port_name(port)); + val |= icl_dpclka_cfgcr0_clk_off(dev_priv, port); + I915_WRITE(DPCLKA_CFGCR0_ICL, val); + } } static void intel_ddi_clk_select(struct intel_encoder *encoder, diff --git a/drivers/gpu/drm/i915/intel_dsi.h b/drivers/gpu/drm/i915/intel_dsi.h index ee93137f4433..d968f1f13e09 100644 --- a/drivers/gpu/drm/i915/intel_dsi.h +++ b/drivers/gpu/drm/i915/intel_dsi.h @@ -146,6 +146,11 @@ static inline bool is_cmd_mode(struct intel_dsi *intel_dsi) return intel_dsi->operation_mode == INTEL_DSI_COMMAND_MODE; } +static inline u16 intel_dsi_encoder_ports(struct intel_encoder *encoder) +{ + return enc_to_intel_dsi(&encoder->base)->ports; +} + /* intel_dsi.c */ int intel_dsi_bitrate(const struct intel_dsi *intel_dsi); int intel_dsi_tlpx_ns(const struct intel_dsi *intel_dsi); From 70a057b7d42582dc846be3d0502dde7e2bd33914 Mon Sep 17 00:00:00 2001 From: Madhav Chauhan Date: Thu, 29 Nov 2018 16:12:18 +0200 Subject: [PATCH 61/98] drm/i915/icl: Calculate DPLL params for DSI This patch calculates various DPLL dividers and parameters for DSI encoder and adjust AFE clock for DSI. For DSI, 8x clock is AFE clock. v2: Extend haswell_crtc_compute_clock() for Gen11 DSI v3: Rebase v4: use port clock instead of bitrate. v5: Reabse and remove divide by 5 v6 by Jani: - Fix indent (Madhav) - Fix dpll state calc for EDP and DP MST Co-developed-by: Vandita Kulkarni Signed-off-by: Madhav Chauhan Signed-off-by: Vandita Kulkarni Signed-off-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/525d41d0d893dcdc8874d2ce70afa226227ea3f4.1543500285.git.jani.nikula@intel.com --- drivers/gpu/drm/i915/intel_display.c | 4 +++- drivers/gpu/drm/i915/intel_dpll_mgr.c | 3 ++- 2 files changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index e349e2661a12..3b4831c2c4b8 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -9333,10 +9333,12 @@ void hsw_disable_pc8(struct drm_i915_private *dev_priv) static int haswell_crtc_compute_clock(struct intel_crtc *crtc, struct intel_crtc_state *crtc_state) { + struct drm_i915_private *dev_priv = to_i915(crtc->base.dev); struct intel_atomic_state *state = to_intel_atomic_state(crtc_state->base.state); - if (!intel_crtc_has_type(crtc_state, INTEL_OUTPUT_DSI)) { + if (!intel_crtc_has_type(crtc_state, INTEL_OUTPUT_DSI) || + IS_ICELAKE(dev_priv)) { struct intel_encoder *encoder = intel_get_crtc_new_encoder(state, crtc_state); diff --git a/drivers/gpu/drm/i915/intel_dpll_mgr.c b/drivers/gpu/drm/i915/intel_dpll_mgr.c index 901e15063b24..d513ca875c67 100644 --- a/drivers/gpu/drm/i915/intel_dpll_mgr.c +++ b/drivers/gpu/drm/i915/intel_dpll_mgr.c @@ -2523,7 +2523,8 @@ static bool icl_calc_dpll_state(struct intel_crtc_state *crtc_state, if (intel_port_is_tc(dev_priv, encoder->port)) ret = icl_calc_tbt_pll(dev_priv, clock, &pll_params); - else if (intel_crtc_has_type(crtc_state, INTEL_OUTPUT_HDMI)) + else if (intel_crtc_has_type(crtc_state, INTEL_OUTPUT_HDMI) || + intel_crtc_has_type(crtc_state, INTEL_OUTPUT_DSI)) ret = cnl_ddi_calculate_wrpll(clock, dev_priv, &pll_params); else ret = icl_calc_dp_combo_pll(dev_priv, clock, &pll_params); From e275804873212fa0585f8d4efee5105ba3e50f83 Mon Sep 17 00:00:00 2001 From: Madhav Chauhan Date: Thu, 29 Nov 2018 16:12:19 +0200 Subject: [PATCH 62/98] drm/i915/icl: Allocate DSI encoder/connector This patch allocates memory for DSI encoder and connector which will be used for various DSI encoder/connector operations and attaching the same to DRM subsystem. This patch also extracts DSI modes info from VBT and save the desired mode info to connector. v2 by Jani: - Drop GEN11 prefix from encoder name - Drop extra parenthesis - Drop extra local variable - Squash encoder power domain here v3 by Jani: - Squash connector and connector helper functions here - Move intel_dsi_vbt_init call here Signed-off-by: Madhav Chauhan Signed-off-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/0197023b92ffa2d59064e30fd4ca22b6a4cff16c.1543500286.git.jani.nikula@intel.com --- drivers/gpu/drm/i915/icl_dsi.c | 117 ++++++++++++++++++++++++++++++--- 1 file changed, 109 insertions(+), 8 deletions(-) diff --git a/drivers/gpu/drm/i915/icl_dsi.c b/drivers/gpu/drm/i915/icl_dsi.c index 01f422df8c23..a40083a7eb6a 100644 --- a/drivers/gpu/drm/i915/icl_dsi.c +++ b/drivers/gpu/drm/i915/icl_dsi.c @@ -26,6 +26,7 @@ */ #include +#include #include "intel_dsi.h" static inline int header_credits_available(struct drm_i915_private *dev_priv, @@ -799,10 +800,9 @@ static void gen11_dsi_powerup_panel(struct intel_encoder *encoder) wait_for_cmds_dispatched_to_panel(encoder); } -static void __attribute__((unused)) -gen11_dsi_pre_enable(struct intel_encoder *encoder, - const struct intel_crtc_state *pipe_config, - const struct drm_connector_state *conn_state) +static void gen11_dsi_pre_enable(struct intel_encoder *encoder, + const struct intel_crtc_state *pipe_config, + const struct drm_connector_state *conn_state) { struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base); @@ -945,10 +945,9 @@ static void gen11_dsi_disable_io_power(struct intel_encoder *encoder) } } -static void __attribute__((unused)) gen11_dsi_disable( - struct intel_encoder *encoder, - const struct intel_crtc_state *old_crtc_state, - const struct drm_connector_state *old_conn_state) +static void gen11_dsi_disable(struct intel_encoder *encoder, + const struct intel_crtc_state *old_crtc_state, + const struct drm_connector_state *old_conn_state) { struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base); @@ -972,10 +971,112 @@ static void __attribute__((unused)) gen11_dsi_disable( gen11_dsi_disable_io_power(encoder); } +static void gen11_dsi_encoder_destroy(struct drm_encoder *encoder) +{ + intel_encoder_destroy(encoder); +} + +static const struct drm_encoder_funcs gen11_dsi_encoder_funcs = { + .destroy = gen11_dsi_encoder_destroy, +}; + +static const struct drm_connector_funcs gen11_dsi_connector_funcs = { + .late_register = intel_connector_register, + .early_unregister = intel_connector_unregister, + .destroy = intel_connector_destroy, + .fill_modes = drm_helper_probe_single_connector_modes, + .atomic_get_property = intel_digital_connector_atomic_get_property, + .atomic_set_property = intel_digital_connector_atomic_set_property, + .atomic_destroy_state = drm_atomic_helper_connector_destroy_state, + .atomic_duplicate_state = intel_digital_connector_duplicate_state, +}; + +static const struct drm_connector_helper_funcs gen11_dsi_connector_helper_funcs = { + .get_modes = intel_dsi_get_modes, + .mode_valid = intel_dsi_mode_valid, + .atomic_check = intel_digital_connector_atomic_check, +}; + void icl_dsi_init(struct drm_i915_private *dev_priv) { + struct drm_device *dev = &dev_priv->drm; + struct intel_dsi *intel_dsi; + struct intel_encoder *encoder; + struct intel_connector *intel_connector; + struct drm_connector *connector; + struct drm_display_mode *scan, *fixed_mode = NULL; enum port port; if (!intel_bios_is_dsi_present(dev_priv, &port)) return; + + intel_dsi = kzalloc(sizeof(*intel_dsi), GFP_KERNEL); + if (!intel_dsi) + return; + + intel_connector = intel_connector_alloc(); + if (!intel_connector) { + kfree(intel_dsi); + return; + } + + encoder = &intel_dsi->base; + intel_dsi->attached_connector = intel_connector; + connector = &intel_connector->base; + + /* register DSI encoder with DRM subsystem */ + drm_encoder_init(dev, &encoder->base, &gen11_dsi_encoder_funcs, + DRM_MODE_ENCODER_DSI, "DSI %c", port_name(port)); + + encoder->pre_enable = gen11_dsi_pre_enable; + encoder->disable = gen11_dsi_disable; + encoder->port = port; + encoder->type = INTEL_OUTPUT_DSI; + encoder->cloneable = 0; + encoder->crtc_mask = BIT(PIPE_A) | BIT(PIPE_B) | BIT(PIPE_C); + encoder->power_domain = POWER_DOMAIN_PORT_DSI; + + /* register DSI connector with DRM subsystem */ + drm_connector_init(dev, connector, &gen11_dsi_connector_funcs, + DRM_MODE_CONNECTOR_DSI); + drm_connector_helper_add(connector, &gen11_dsi_connector_helper_funcs); + connector->display_info.subpixel_order = SubPixelHorizontalRGB; + connector->interlace_allowed = false; + connector->doublescan_allowed = false; + + /* attach connector to encoder */ + intel_connector_attach_encoder(intel_connector, encoder); + + /* fill mode info from VBT */ + mutex_lock(&dev->mode_config.mutex); + intel_dsi_vbt_get_modes(intel_dsi); + list_for_each_entry(scan, &connector->probed_modes, head) { + if (scan->type & DRM_MODE_TYPE_PREFERRED) { + fixed_mode = drm_mode_duplicate(dev, scan); + break; + } + } + mutex_unlock(&dev->mode_config.mutex); + + if (!fixed_mode) { + DRM_ERROR("DSI fixed mode info missing\n"); + goto err; + } + + connector->display_info.width_mm = fixed_mode->width_mm; + connector->display_info.height_mm = fixed_mode->height_mm; + intel_panel_init(&intel_connector->panel, fixed_mode, NULL); + intel_panel_setup_backlight(connector, INVALID_PIPE); + + if (!intel_dsi_vbt_init(intel_dsi, MIPI_DSI_GENERIC_PANEL_ID)) { + DRM_DEBUG_KMS("no device found\n"); + goto err; + } + + return; + +err: + drm_encoder_cleanup(&encoder->base); + kfree(intel_dsi); + kfree(intel_connector); } From 95f2f4dbb19ff097fc86f4ea7529e6453bab3eb9 Mon Sep 17 00:00:00 2001 From: Vandita Kulkarni Date: Thu, 29 Nov 2018 16:12:20 +0200 Subject: [PATCH 63/98] drm/i915/icl: Use the same pll functions for dsi The same pll manager functions can be used to enable dpll for mipi. Hence enabling the IO power and esc clock as part of pre pll enable call. v2 by Jani: - fix function parameter indent (Madhav) Signed-off-by: Vandita Kulkarni Reviewed-by: Madhav Chauhan Signed-off-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/07db5b1ce59c00295fad8a8e81908d59d8d76915.1543500286.git.jani.nikula@intel.com --- drivers/gpu/drm/i915/icl_dsi.c | 16 +++++++++++----- 1 file changed, 11 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/i915/icl_dsi.c b/drivers/gpu/drm/i915/icl_dsi.c index a40083a7eb6a..9d6cbe35259a 100644 --- a/drivers/gpu/drm/i915/icl_dsi.c +++ b/drivers/gpu/drm/i915/icl_dsi.c @@ -800,17 +800,22 @@ static void gen11_dsi_powerup_panel(struct intel_encoder *encoder) wait_for_cmds_dispatched_to_panel(encoder); } -static void gen11_dsi_pre_enable(struct intel_encoder *encoder, - const struct intel_crtc_state *pipe_config, - const struct drm_connector_state *conn_state) +static void gen11_dsi_pre_pll_enable(struct intel_encoder *encoder, + const struct intel_crtc_state *pipe_config, + const struct drm_connector_state *conn_state) { - struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base); - /* step2: enable IO power */ gen11_dsi_enable_io_power(encoder); /* step3: enable DSI PLL */ gen11_dsi_program_esc_clk_div(encoder); +} + +static void gen11_dsi_pre_enable(struct intel_encoder *encoder, + const struct intel_crtc_state *pipe_config, + const struct drm_connector_state *conn_state) +{ + struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base); /* step4: enable DSI port and DPHY */ gen11_dsi_enable_port_and_phy(encoder, pipe_config); @@ -1028,6 +1033,7 @@ void icl_dsi_init(struct drm_i915_private *dev_priv) drm_encoder_init(dev, &encoder->base, &gen11_dsi_encoder_funcs, DRM_MODE_ENCODER_DSI, "DSI %c", port_name(port)); + encoder->pre_pll_enable = gen11_dsi_pre_pll_enable; encoder->pre_enable = gen11_dsi_pre_enable; encoder->disable = gen11_dsi_disable; encoder->port = port; From 972d607c59ed4170cd25350d22ded19477360cb9 Mon Sep 17 00:00:00 2001 From: Madhav Chauhan Date: Thu, 29 Nov 2018 16:12:21 +0200 Subject: [PATCH 64/98] drm/i915/icl: Fill DSI ports info This patch fills backlight, CABC and general port info for Gen11 DSI. Signed-off-by: Madhav Chauhan Reviewed-by: Jani Nikula Signed-off-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/48c9f867d61a60dc7c2ce744ac2325655652c55f.1543500286.git.jani.nikula@intel.com --- drivers/gpu/drm/i915/icl_dsi.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/drivers/gpu/drm/i915/icl_dsi.c b/drivers/gpu/drm/i915/icl_dsi.c index 9d6cbe35259a..dc3fcc67814b 100644 --- a/drivers/gpu/drm/i915/icl_dsi.c +++ b/drivers/gpu/drm/i915/icl_dsi.c @@ -1074,6 +1074,14 @@ void icl_dsi_init(struct drm_i915_private *dev_priv) intel_panel_init(&intel_connector->panel, fixed_mode, NULL); intel_panel_setup_backlight(connector, INVALID_PIPE); + if (dev_priv->vbt.dsi.config->dual_link) + intel_dsi->ports = BIT(PORT_A) | BIT(PORT_B); + else + intel_dsi->ports = BIT(port); + + intel_dsi->dcs_backlight_ports = dev_priv->vbt.dsi.bl_ports; + intel_dsi->dcs_cabc_ports = dev_priv->vbt.dsi.cabc_ports; + if (!intel_dsi_vbt_init(intel_dsi, MIPI_DSI_GENERIC_PANEL_ID)) { DRM_DEBUG_KMS("no device found\n"); goto err; From c5f9c934936e33198f8c43e0a7c261aac5960344 Mon Sep 17 00:00:00 2001 From: Madhav Chauhan Date: Thu, 29 Nov 2018 16:12:22 +0200 Subject: [PATCH 65/98] drm/i915/icl: Allocate DSI hosts and imlement host transfer Allocate DSI host structure for each DSI port available on gen11 and register them with DSI fwk of DRM. Some of the DSI host operations are also registered as part of this. Retrieves DSI pkt (from DSI msg) to be sent over DSI link using DRM DSI exported functions. A wrapper function is also added as "DSI host transfer" for sending DSI data/cmd. Add DSI packet payload to command payload queue using credit based mechanism for *long* packets. v2 by Jani: - indentation - Use the new credit available helper - Use int for free_credits - Add intel_dsi local variable for better code flow - Use the new credit available helper - Use int for free_credits, i, and j v3 by Jani: - Squash DSI host allocation and transfer patches together Signed-off-by: Madhav Chauhan Signed-off-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/af4f168ed8737d44687d8b6f21ecaa7e805eb695.1543500286.git.jani.nikula@intel.com --- drivers/gpu/drm/i915/icl_dsi.c | 147 +++++++++++++++++++++++++++++++++ 1 file changed, 147 insertions(+) diff --git a/drivers/gpu/drm/i915/icl_dsi.c b/drivers/gpu/drm/i915/icl_dsi.c index dc3fcc67814b..43a4d19e36d8 100644 --- a/drivers/gpu/drm/i915/icl_dsi.c +++ b/drivers/gpu/drm/i915/icl_dsi.c @@ -108,6 +108,90 @@ static void wait_for_cmds_dispatched_to_panel(struct intel_encoder *encoder) } } +static bool add_payld_to_queue(struct intel_dsi_host *host, const u8 *data, + u32 len) +{ + struct intel_dsi *intel_dsi = host->intel_dsi; + struct drm_i915_private *dev_priv = to_i915(intel_dsi->base.base.dev); + enum transcoder dsi_trans = dsi_port_to_transcoder(host->port); + int free_credits; + int i, j; + + for (i = 0; i < len; i += 4) { + u32 tmp = 0; + + free_credits = payload_credits_available(dev_priv, dsi_trans); + if (free_credits < 1) { + DRM_ERROR("Payload credit not available\n"); + return false; + } + + for (j = 0; j < min_t(u32, len - i, 4); j++) + tmp |= *data++ << 8 * j; + + I915_WRITE(DSI_CMD_TXPYLD(dsi_trans), tmp); + } + + return true; +} + +static int dsi_send_pkt_hdr(struct intel_dsi_host *host, + struct mipi_dsi_packet pkt, bool enable_lpdt) +{ + struct intel_dsi *intel_dsi = host->intel_dsi; + struct drm_i915_private *dev_priv = to_i915(intel_dsi->base.base.dev); + enum transcoder dsi_trans = dsi_port_to_transcoder(host->port); + u32 tmp; + int free_credits; + + /* check if header credit available */ + free_credits = header_credits_available(dev_priv, dsi_trans); + if (free_credits < 1) { + DRM_ERROR("send pkt header failed, not enough hdr credits\n"); + return -1; + } + + tmp = I915_READ(DSI_CMD_TXHDR(dsi_trans)); + + if (pkt.payload) + tmp |= PAYLOAD_PRESENT; + else + tmp &= ~PAYLOAD_PRESENT; + + tmp &= ~VBLANK_FENCE; + + if (enable_lpdt) + tmp |= LP_DATA_TRANSFER; + + tmp &= ~(PARAM_WC_MASK | VC_MASK | DT_MASK); + tmp |= ((pkt.header[0] & VC_MASK) << VC_SHIFT); + tmp |= ((pkt.header[0] & DT_MASK) << DT_SHIFT); + tmp |= (pkt.header[1] << PARAM_WC_LOWER_SHIFT); + tmp |= (pkt.header[2] << PARAM_WC_UPPER_SHIFT); + I915_WRITE(DSI_CMD_TXHDR(dsi_trans), tmp); + + return 0; +} + +static int dsi_send_pkt_payld(struct intel_dsi_host *host, + struct mipi_dsi_packet pkt) +{ + /* payload queue can accept *256 bytes*, check limit */ + if (pkt.payload_length > MAX_PLOAD_CREDIT * 4) { + DRM_ERROR("payload size exceeds max queue limit\n"); + return -1; + } + + /* load data into command payload queue */ + if (!add_payld_to_queue(host, pkt.payload, + pkt.payload_length)) { + DRM_ERROR("adding payload to queue failed\n"); + return -1; + } + + return 0; +} + static void dsi_program_swing_and_deemphasis(struct intel_encoder *encoder) { struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); @@ -1002,6 +1086,58 @@ static const struct drm_connector_helper_funcs gen11_dsi_connector_helper_funcs .atomic_check = intel_digital_connector_atomic_check, }; +static int gen11_dsi_host_attach(struct mipi_dsi_host *host, + struct mipi_dsi_device *dsi) +{ + return 0; +} + +static int gen11_dsi_host_detach(struct mipi_dsi_host *host, + struct mipi_dsi_device *dsi) +{ + return 0; +} + +static ssize_t gen11_dsi_host_transfer(struct mipi_dsi_host *host, + const struct mipi_dsi_msg *msg) +{ + struct intel_dsi_host *intel_dsi_host = to_intel_dsi_host(host); + struct mipi_dsi_packet dsi_pkt; + ssize_t ret; + bool enable_lpdt = false; + + ret = mipi_dsi_create_packet(&dsi_pkt, msg); + if (ret < 0) + return ret; + + if (msg->flags & MIPI_DSI_MSG_USE_LPM) + enable_lpdt = true; + + /* send packet header */ + ret = dsi_send_pkt_hdr(intel_dsi_host, dsi_pkt, enable_lpdt); + if (ret < 0) + return ret; + + /* only long packet contains payload */ + if (mipi_dsi_packet_format_is_long(msg->type)) { + ret = dsi_send_pkt_payld(intel_dsi_host, dsi_pkt); + if (ret < 0) + return ret; + } + + //TODO: add payload receive code if needed + + ret = sizeof(dsi_pkt.header) + dsi_pkt.payload_length; + + return ret; +} + +static const struct mipi_dsi_host_ops gen11_dsi_host_ops = { + .attach = gen11_dsi_host_attach, + .detach = gen11_dsi_host_detach, + .transfer = gen11_dsi_host_transfer, +}; + void icl_dsi_init(struct drm_i915_private *dev_priv) { struct drm_device *dev = &dev_priv->drm; @@ -1074,6 +1210,7 @@ void icl_dsi_init(struct drm_i915_private *dev_priv) intel_panel_init(&intel_connector->panel, fixed_mode, NULL); intel_panel_setup_backlight(connector, INVALID_PIPE); + if (dev_priv->vbt.dsi.config->dual_link) intel_dsi->ports = BIT(PORT_A) | BIT(PORT_B); else @@ -1082,6 +1219,16 @@ void icl_dsi_init(struct drm_i915_private *dev_priv) intel_dsi->dcs_backlight_ports = dev_priv->vbt.dsi.bl_ports; intel_dsi->dcs_cabc_ports = dev_priv->vbt.dsi.cabc_ports; + for_each_dsi_port(port, intel_dsi->ports) { + struct intel_dsi_host *host; + + host = intel_dsi_host_init(intel_dsi, &gen11_dsi_host_ops, port); + if (!host) + goto err; + + intel_dsi->dsi_hosts[port] = host; + } + if (!intel_dsi_vbt_init(intel_dsi, MIPI_DSI_GENERIC_PANEL_ID)) { DRM_DEBUG_KMS("no device found\n"); goto err; From 8327af281d292524822bae5388d76742750e63b5 Mon Sep 17 00:00:00 2001 From: Vandita Kulkarni Date: Thu, 29 Nov 2018 16:12:23 +0200 Subject: [PATCH 66/98] drm/i915/icl: Add get config functionality for DSI This patch implements the functionality for getting PIPE configuration to which DSI encoder is connected. Use the same method to get port clock like other DDI encoders. Used during the atomic modeset. v2 by Jani: - Squash Madhav's and Vandita's get config bits together - Move cnl_calc_wrpll_link() to intel_drv.h - Drop extra temp variables - Use enc_to_intel_dsi() instead of open coding Co-developed-by: Madhav Chauhan Signed-off-by: Vandita Kulkarni Signed-off-by: Madhav Chauhan Signed-off-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/f21fa4258068d04582f2bf30735e5536a8043bdf.1543500286.git.jani.nikula@intel.com --- drivers/gpu/drm/i915/icl_dsi.c | 15 +++++++++++++++ drivers/gpu/drm/i915/intel_ddi.c | 4 ++-- drivers/gpu/drm/i915/intel_drv.h | 2 ++ 3 files changed, 19 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/icl_dsi.c b/drivers/gpu/drm/i915/icl_dsi.c index 43a4d19e36d8..fba68e1e1bd4 100644 --- a/drivers/gpu/drm/i915/icl_dsi.c +++ b/drivers/gpu/drm/i915/icl_dsi.c @@ -1060,6 +1060,20 @@ static void gen11_dsi_disable(struct intel_encoder *encoder, gen11_dsi_disable_io_power(encoder); } +static void gen11_dsi_get_config(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config) +{ + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); + struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base); + u32 pll_id; + + /* FIXME: adapt icl_ddi_clock_get() for DSI and use that? */ + pll_id = intel_get_shared_dpll_id(dev_priv, pipe_config->shared_dpll); + pipe_config->port_clock = cnl_calc_wrpll_link(dev_priv, pll_id); + pipe_config->base.adjusted_mode.crtc_clock = intel_dsi->pclk; + pipe_config->output_types |= BIT(INTEL_OUTPUT_DSI); +} + static void gen11_dsi_encoder_destroy(struct drm_encoder *encoder) { intel_encoder_destroy(encoder); @@ -1173,6 +1187,7 @@ void icl_dsi_init(struct drm_i915_private *dev_priv) encoder->pre_enable = gen11_dsi_pre_enable; encoder->disable = gen11_dsi_disable; encoder->port = port; + encoder->get_config = gen11_dsi_get_config; encoder->type = INTEL_OUTPUT_DSI; encoder->cloneable = 0; encoder->crtc_mask = BIT(PIPE_A) | BIT(PIPE_B) | BIT(PIPE_C); diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c index 6308d0f151e1..f3e1d6a0b7dd 100644 --- a/drivers/gpu/drm/i915/intel_ddi.c +++ b/drivers/gpu/drm/i915/intel_ddi.c @@ -1364,8 +1364,8 @@ static int skl_calc_wrpll_link(struct drm_i915_private *dev_priv, return dco_freq / (p0 * p1 * p2 * 5); } -static int cnl_calc_wrpll_link(struct drm_i915_private *dev_priv, - enum intel_dpll_id pll_id) +int cnl_calc_wrpll_link(struct drm_i915_private *dev_priv, + enum intel_dpll_id pll_id) { uint32_t cfgcr0, cfgcr1; uint32_t p0, p1, p2, dco_freq, ref_clock; diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index 04b9545994f2..f94a04b4ad87 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -1535,6 +1535,8 @@ u8 intel_ddi_dp_pre_emphasis_max(struct intel_encoder *encoder, int intel_ddi_toggle_hdcp_signalling(struct intel_encoder *intel_encoder, bool enable); void icl_sanitize_encoder_pll_mapping(struct intel_encoder *encoder); +int cnl_calc_wrpll_link(struct drm_i915_private *dev_priv, + enum intel_dpll_id pll_id); unsigned int intel_fb_align_height(const struct drm_framebuffer *fb, int color_plane, unsigned int height); From ab8411483a3e37fd104c9947ead5d687aa1d1ad9 Mon Sep 17 00:00:00 2001 From: Madhav Chauhan Date: Thu, 29 Nov 2018 16:12:24 +0200 Subject: [PATCH 67/98] drm/i915/icl: Get HW state for DSI encoder This patch read out the current hw state for DSI and return true if encoder is active. v2 by Jani: - Squash connector get hw state hook here - Squash encode get hw state fix here v3 by Jani: - Add encoder->get_power_domains() (Imre) v4 by Jani: - Make encoder->get_power_domains() sensible... (Imre) v5 by Jani: - Power domains are bit positions, not bits (Stan, Imre) Cc: Imre Deak Signed-off-by: Madhav Chauhan Signed-off-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/ec6da32a55b9fb045527f14e41ed3dce86d46a97.1543500286.git.jani.nikula@intel.com --- drivers/gpu/drm/i915/icl_dsi.c | 59 ++++++++++++++++++++++++++++++++++ 1 file changed, 59 insertions(+) diff --git a/drivers/gpu/drm/i915/icl_dsi.c b/drivers/gpu/drm/i915/icl_dsi.c index fba68e1e1bd4..12e907b00bb9 100644 --- a/drivers/gpu/drm/i915/icl_dsi.c +++ b/drivers/gpu/drm/i915/icl_dsi.c @@ -1074,6 +1074,62 @@ static void gen11_dsi_get_config(struct intel_encoder *encoder, pipe_config->output_types |= BIT(INTEL_OUTPUT_DSI); } +static u64 gen11_dsi_get_power_domains(struct intel_encoder *encoder, + struct intel_crtc_state *crtc_state) +{ + struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base); + u64 domains = 0; + enum port port; + + for_each_dsi_port(port, intel_dsi->ports) + if (port == PORT_A) + domains |= BIT_ULL(POWER_DOMAIN_PORT_DDI_A_IO); + else + domains |= BIT_ULL(POWER_DOMAIN_PORT_DDI_B_IO); + + return domains; +} + +static bool gen11_dsi_get_hw_state(struct intel_encoder *encoder, + enum pipe *pipe) +{ + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); + struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base); + u32 tmp; + enum port port; + enum transcoder dsi_trans; + bool ret = false; + + if (!intel_display_power_get_if_enabled(dev_priv, + encoder->power_domain)) + return false; + + for_each_dsi_port(port, intel_dsi->ports) { + dsi_trans = dsi_port_to_transcoder(port); + tmp = I915_READ(TRANS_DDI_FUNC_CTL(dsi_trans)); + switch (tmp & TRANS_DDI_EDP_INPUT_MASK) { + case TRANS_DDI_EDP_INPUT_A_ON: + *pipe = PIPE_A; + break; + case TRANS_DDI_EDP_INPUT_B_ONOFF: + *pipe = PIPE_B; + break; + case TRANS_DDI_EDP_INPUT_C_ONOFF: + *pipe = PIPE_C; + break; + default: + DRM_ERROR("Invalid PIPE input\n"); + goto out; + } + + tmp = I915_READ(PIPECONF(dsi_trans)); + ret = tmp & PIPECONF_ENABLE; + } +out: + intel_display_power_put(dev_priv, encoder->power_domain); + return ret; +} + static void gen11_dsi_encoder_destroy(struct drm_encoder *encoder) { intel_encoder_destroy(encoder); @@ -1188,10 +1244,12 @@ void icl_dsi_init(struct drm_i915_private *dev_priv) encoder->disable = gen11_dsi_disable; encoder->port = port; encoder->get_config = gen11_dsi_get_config; + encoder->get_hw_state = gen11_dsi_get_hw_state; encoder->type = INTEL_OUTPUT_DSI; encoder->cloneable = 0; encoder->crtc_mask = BIT(PIPE_A) | BIT(PIPE_B) | BIT(PIPE_C); encoder->power_domain = POWER_DOMAIN_PORT_DSI; + encoder->get_power_domains = gen11_dsi_get_power_domains; /* register DSI connector with DRM subsystem */ drm_connector_init(dev, connector, &gen11_dsi_connector_funcs, @@ -1200,6 +1258,7 @@ void icl_dsi_init(struct drm_i915_private *dev_priv) connector->display_info.subpixel_order = SubPixelHorizontalRGB; connector->interlace_allowed = false; connector->doublescan_allowed = false; + intel_connector->get_hw_state = intel_connector_get_hw_state; /* attach connector to encoder */ intel_connector_attach_encoder(intel_connector, encoder); From d04afb150172d0bd96bf115b62a6208bdc7e01ce Mon Sep 17 00:00:00 2001 From: Madhav Chauhan Date: Thu, 29 Nov 2018 16:12:25 +0200 Subject: [PATCH 68/98] drm/i915/icl: Add DSI encoder compute config hook This patch implements compute config for Gen11 DSI encoder which is required at the time of modeset. For DSI 8X clock is AFE clock which is 5 times port clock. v2 by Jani: - drop the enable nop hook - fixed_mode is always true - HAS_GMCH_DISPLAY() is always false v3 by Jani: - set encoder->compute_config dropped during rebase v4 by Jani: - squash Vandita's port clock patch - remove todo comment Co-developed-by: Vandita Kulkarni Signed-off-by: Madhav Chauhan Signed-off-by: Vandita Kulkarni Signed-off-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/a21574173caa5e2932d9e3c537b0931097ab5ac2.1543500286.git.jani.nikula@intel.com --- drivers/gpu/drm/i915/icl_dsi.c | 31 +++++++++++++++++++++++++++++++ 1 file changed, 31 insertions(+) diff --git a/drivers/gpu/drm/i915/icl_dsi.c b/drivers/gpu/drm/i915/icl_dsi.c index 12e907b00bb9..f935008ff960 100644 --- a/drivers/gpu/drm/i915/icl_dsi.c +++ b/drivers/gpu/drm/i915/icl_dsi.c @@ -1074,6 +1074,36 @@ static void gen11_dsi_get_config(struct intel_encoder *encoder, pipe_config->output_types |= BIT(INTEL_OUTPUT_DSI); } +static bool gen11_dsi_compute_config(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) +{ + struct intel_dsi *intel_dsi = container_of(encoder, struct intel_dsi, + base); + struct intel_connector *intel_connector = intel_dsi->attached_connector; + struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc); + const struct drm_display_mode *fixed_mode = + intel_connector->panel.fixed_mode; + struct drm_display_mode *adjusted_mode = + &pipe_config->base.adjusted_mode; + + intel_fixed_panel_mode(fixed_mode, adjusted_mode); + intel_pch_panel_fitting(crtc, pipe_config, conn_state->scaling_mode); + + adjusted_mode->flags = 0; + + /* Dual link goes to trancoder DSI'0' */ + if (intel_dsi->ports == BIT(PORT_B)) + pipe_config->cpu_transcoder = TRANSCODER_DSI_1; + else + pipe_config->cpu_transcoder = TRANSCODER_DSI_0; + + pipe_config->clock_set = true; + pipe_config->port_clock = intel_dsi_bitrate(intel_dsi) / 5; + + return true; +} + static u64 gen11_dsi_get_power_domains(struct intel_encoder *encoder, struct intel_crtc_state *crtc_state) { @@ -1244,6 +1274,7 @@ void icl_dsi_init(struct drm_i915_private *dev_priv) encoder->disable = gen11_dsi_disable; encoder->port = port; encoder->get_config = gen11_dsi_get_config; + encoder->compute_config = gen11_dsi_compute_config; encoder->get_hw_state = gen11_dsi_get_hw_state; encoder->type = INTEL_OUTPUT_DSI; encoder->cloneable = 0; From 5a8507b5aa84fe26f2511fd667aa5965499c9d82 Mon Sep 17 00:00:00 2001 From: Madhav Chauhan Date: Thu, 29 Nov 2018 16:12:26 +0200 Subject: [PATCH 69/98] drm/i915/icl: Configure DSI Dual link mode This patch configures DSI video mode dual link by programming DSS_CTL registers. v2: Use new bitfield definitions from Anusha's patch Correct register to be programmed and use max depth buffer value (James) v3 by Jani: - checkpatch fixes Signed-off-by: Madhav Chauhan Signed-off-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/d8962f4e1beffc2099b8bef8cb5a5191aa1d9efd.1543500286.git.jani.nikula@intel.com --- drivers/gpu/drm/i915/icl_dsi.c | 42 +++++++++++++++++++++++++++++++++- 1 file changed, 41 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/icl_dsi.c b/drivers/gpu/drm/i915/icl_dsi.c index f935008ff960..2b576a70222f 100644 --- a/drivers/gpu/drm/i915/icl_dsi.c +++ b/drivers/gpu/drm/i915/icl_dsi.c @@ -257,6 +257,45 @@ static void dsi_program_swing_and_deemphasis(struct intel_encoder *encoder) } } +static void configure_dual_link_mode(struct intel_encoder *encoder, + const struct intel_crtc_state *pipe_config) +{ + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); + struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base); + u32 dss_ctl1; + + dss_ctl1 = I915_READ(DSS_CTL1); + dss_ctl1 |= SPLITTER_ENABLE; + dss_ctl1 &= ~OVERLAP_PIXELS_MASK; + dss_ctl1 |= OVERLAP_PIXELS(intel_dsi->pixel_overlap); + + if (intel_dsi->dual_link == DSI_DUAL_LINK_FRONT_BACK) { + const struct drm_display_mode *adjusted_mode = + &pipe_config->base.adjusted_mode; + u32 dss_ctl2; + u16 hactive = adjusted_mode->crtc_hdisplay; + u16 dl_buffer_depth; + + dss_ctl1 &= ~DUAL_LINK_MODE_INTERLEAVE; + dl_buffer_depth = hactive / 2 + intel_dsi->pixel_overlap; + + if (dl_buffer_depth > MAX_DL_BUFFER_TARGET_DEPTH) + DRM_ERROR("DL buffer depth exceed max value\n"); + + dss_ctl1 &= ~LEFT_DL_BUF_TARGET_DEPTH_MASK; + dss_ctl1 |= LEFT_DL_BUF_TARGET_DEPTH(dl_buffer_depth); + dss_ctl2 = I915_READ(DSS_CTL2); + dss_ctl2 &= ~RIGHT_DL_BUF_TARGET_DEPTH_MASK; + dss_ctl2 |= RIGHT_DL_BUF_TARGET_DEPTH(dl_buffer_depth); + I915_WRITE(DSS_CTL2, dss_ctl2); + } else { + /* Interleave */ + dss_ctl1 |= DUAL_LINK_MODE_INTERLEAVE; + } + + I915_WRITE(DSS_CTL1, dss_ctl1); +} + static void gen11_dsi_program_esc_clk_div(struct intel_encoder *encoder) { struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); @@ -591,7 +630,8 @@ gen11_dsi_configure_transcoder(struct intel_encoder *encoder, I915_WRITE(TRANS_DDI_FUNC_CTL2(dsi_trans), tmp); } - //TODO: configure DSS_CTL1 + /* configure stream splitting */ + configure_dual_link_mode(encoder, pipe_config); } for_each_dsi_port(port, intel_dsi->ports) { From 2ca711caeca2c6a4f3026d9fbdb135b65d7d68b3 Mon Sep 17 00:00:00 2001 From: Madhav Chauhan Date: Thu, 29 Nov 2018 16:12:27 +0200 Subject: [PATCH 70/98] drm/i915/icl: Consider DSI for getting transcoder state For Gen11 DSI, we use similar registers like for eDP to find if DSI encoder is connected or not to a pipe. This patch refactors existing hsw_get_transcoder_state() to handle this. v2 by Jani: - Add WARN_ON(dsi && edp) (Ville) Signed-off-by: Madhav Chauhan Signed-off-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/2e10b977dfc7aa985a8559d6cd59ed0981848e95.1543500286.git.jani.nikula@intel.com --- drivers/gpu/drm/i915/intel_display.c | 31 +++++++++++++++++++++------- 1 file changed, 24 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 3b4831c2c4b8..9b5753cdc76b 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -9477,6 +9477,8 @@ static bool hsw_get_transcoder_state(struct intel_crtc *crtc, struct drm_i915_private *dev_priv = to_i915(dev); enum intel_display_power_domain power_domain; u32 tmp; + bool is_dsi = false; + bool is_edp = false; /* * The pipe->transcoder mapping is fixed with the exception of the eDP @@ -9489,26 +9491,41 @@ static bool hsw_get_transcoder_state(struct intel_crtc *crtc, * consistency and less surprising code; it's in always on power). */ tmp = I915_READ(TRANS_DDI_FUNC_CTL(TRANSCODER_EDP)); - if (tmp & TRANS_DDI_FUNC_ENABLE) { - enum pipe trans_edp_pipe; + if (tmp & TRANS_DDI_FUNC_ENABLE) + is_edp = true; + + if (IS_ICELAKE(dev_priv)) { + tmp = I915_READ(TRANS_DDI_FUNC_CTL(TRANSCODER_DSI_0)); + if (tmp & TRANS_DDI_FUNC_ENABLE) + is_dsi = true; + } + + WARN_ON(is_edp && is_dsi); + + if (is_edp || is_dsi) { + enum pipe trans_pipe; switch (tmp & TRANS_DDI_EDP_INPUT_MASK) { default: WARN(1, "unknown pipe linked to edp transcoder\n"); /* fall through */ case TRANS_DDI_EDP_INPUT_A_ONOFF: case TRANS_DDI_EDP_INPUT_A_ON: - trans_edp_pipe = PIPE_A; + trans_pipe = PIPE_A; break; case TRANS_DDI_EDP_INPUT_B_ONOFF: - trans_edp_pipe = PIPE_B; + trans_pipe = PIPE_B; break; case TRANS_DDI_EDP_INPUT_C_ONOFF: - trans_edp_pipe = PIPE_C; + trans_pipe = PIPE_C; break; } - if (trans_edp_pipe == crtc->pipe) - pipe_config->cpu_transcoder = TRANSCODER_EDP; + if (trans_pipe == crtc->pipe) { + if (is_edp) + pipe_config->cpu_transcoder = TRANSCODER_EDP; + else if (is_dsi) + pipe_config->cpu_transcoder = TRANSCODER_DSI_0; + } } power_domain = POWER_DOMAIN_TRANSCODER(pipe_config->cpu_transcoder); From 2eae5d6bfa5f5d3e815cd0c76d864c0ff09b2c4c Mon Sep 17 00:00:00 2001 From: Madhav Chauhan Date: Thu, 29 Nov 2018 16:12:28 +0200 Subject: [PATCH 71/98] drm/i915/icl: Get pipe timings for DSI Transcoder timings for Gen11 DSI encoder is available at pipe level unlike in older platform where port specific registers need to be accessed. v2 by Jani: - get timings for (!dsi || icl) instead of (dsi && icl). Signed-off-by: Madhav Chauhan Signed-off-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/f60e0c1aee08248e758da3219d3239898b43ba41.1543500286.git.jani.nikula@intel.com --- drivers/gpu/drm/i915/intel_display.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 9b5753cdc76b..58cadd251acc 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -9660,7 +9660,8 @@ static bool haswell_get_pipe_config(struct intel_crtc *crtc, if (!active) goto out; - if (!transcoder_is_dsi(pipe_config->cpu_transcoder)) { + if (!transcoder_is_dsi(pipe_config->cpu_transcoder) || + IS_ICELAKE(dev_priv)) { haswell_get_ddi_port_state(crtc, pipe_config); intel_get_pipe_timings(crtc, pipe_config); } From 05f2f03dd206177fda22d9e442e8becaaf502093 Mon Sep 17 00:00:00 2001 From: Madhav Chauhan Date: Thu, 29 Nov 2018 16:12:29 +0200 Subject: [PATCH 72/98] drm/i915/icl: Define missing bitfield for shortplug reg This patch define missing bitfield for shortplug ctl ddi register which will be used for ICL DSI GPIO programming. Signed-off-by: Madhav Chauhan Signed-off-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/4b9feb75eb6c303556b91c8a23c505a4593a99a1.1543500286.git.jani.nikula@intel.com --- drivers/gpu/drm/i915/i915_reg.h | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index d3ef97915455..776f445fca25 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -7751,6 +7751,7 @@ enum { #define ICP_DDIB_HPD_LONG_DETECT (2 << 4) #define ICP_DDIB_HPD_SHORT_LONG_DETECT (3 << 4) #define ICP_DDIA_HPD_ENABLE (1 << 3) +#define ICP_DDIA_HPD_OP_DRIVE_1 (1 << 2) #define ICP_DDIA_HPD_STATUS_MASK (3 << 0) #define ICP_DDIA_HPD_NO_DETECT (0 << 0) #define ICP_DDIA_HPD_SHORT_DETECT (1 << 0) From f4ff2120301d85ce18438dbb64ad4bf5e9382506 Mon Sep 17 00:00:00 2001 From: Madhav Chauhan Date: Thu, 29 Nov 2018 16:12:30 +0200 Subject: [PATCH 73/98] drm/i915/icl: Define Panel power ctrl register There are two panel power sequencers. Each register has two addressable instances. This patch defines both the instances of Panel power control register Signed-off-by: Madhav Chauhan Signed-off-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/99bb687b17a9165527a6210a79271c8175c8a4e3.1543500286.git.jani.nikula@intel.com --- drivers/gpu/drm/i915/i915_reg.h | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index 776f445fca25..0a7d60509ca7 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -4618,6 +4618,17 @@ enum { #define _PP_STATUS 0x61200 #define PP_STATUS(pps_idx) _MMIO_PPS(pps_idx, _PP_STATUS) #define PP_ON (1 << 31) + +#define _PP_CONTROL_1 0xc7204 +#define _PP_CONTROL_2 0xc7304 +#define ICP_PP_CONTROL(x) _MMIO(((x) == 1) ? _PP_CONTROL_1 : \ + _PP_CONTROL_2) +#define POWER_CYCLE_DELAY_MASK (0x1f << 4) +#define POWER_CYCLE_DELAY_SHIFT 4 +#define VDD_OVERRIDE_FORCE (1 << 3) +#define BACKLIGHT_ENABLE (1 << 2) +#define PWR_DOWN_ON_RESET (1 << 1) +#define PWR_STATE_TARGET (1 << 0) /* * Indicates that all dependencies of the panel are on: * From 56b7b1aa13b971e50402a63c37aaf244644b4f8a Mon Sep 17 00:00:00 2001 From: Madhav Chauhan Date: Thu, 29 Nov 2018 16:12:31 +0200 Subject: [PATCH 74/98] drm/i915/icl: Define display GPIO pins for DSI Display Pins are the only GPIOs that need to be used by driver for DSI panels. So driver should now have its own implementation to toggle these pins based on GPIO info received from VBT sequences. Signed-off-by: Madhav Chauhan Signed-off-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/e7e3bb05d4f48b1876169a69f495bcf6d511fda5.1543500286.git.jani.nikula@intel.com --- drivers/gpu/drm/i915/intel_dsi_vbt.c | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/drivers/gpu/drm/i915/intel_dsi_vbt.c b/drivers/gpu/drm/i915/intel_dsi_vbt.c index a72de81f4832..b41ca6436401 100644 --- a/drivers/gpu/drm/i915/intel_dsi_vbt.c +++ b/drivers/gpu/drm/i915/intel_dsi_vbt.c @@ -103,6 +103,18 @@ static struct gpio_map vlv_gpio_table[] = { #define CHV_GPIO_PAD_CFG1(f, i) (0x4400 + (f) * 0x400 + (i) * 8 + 4) #define CHV_GPIO_CFGLOCK (1 << 31) +/* ICL DSI Display GPIO Pins */ +#define ICL_GPIO_DDSP_HPD_A 0 +#define ICL_GPIO_L_VDDEN_1 1 +#define ICL_GPIO_L_BKLTEN_1 2 +#define ICL_GPIO_DDPA_CTRLCLK_1 3 +#define ICL_GPIO_DDPA_CTRLDATA_1 4 +#define ICL_GPIO_DDSP_HPD_B 5 +#define ICL_GPIO_L_VDDEN_2 6 +#define ICL_GPIO_L_BKLTEN_2 7 +#define ICL_GPIO_DDPA_CTRLCLK_2 8 +#define ICL_GPIO_DDPA_CTRLDATA_2 9 + static inline enum port intel_dsi_seq_port_to_port(u8 port) { return port ? PORT_C : PORT_A; From 690c318ed8e06c2927c11fae366e55277f0cb42a Mon Sep 17 00:00:00 2001 From: Jani Nikula Date: Thu, 29 Nov 2018 16:12:32 +0200 Subject: [PATCH 75/98] drm/i915/icl: add dummy DSI GPIO element execution function Add dummy debug logging GPIO element execution function for ICL. Reviewed-by: Madhav Chauhan Signed-off-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/f59009ea0a0ebc489a5ec66f387d9dcf7264141f.1543500286.git.jani.nikula@intel.com --- drivers/gpu/drm/i915/intel_dsi_vbt.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/intel_dsi_vbt.c b/drivers/gpu/drm/i915/intel_dsi_vbt.c index b41ca6436401..a1a8b3790e61 100644 --- a/drivers/gpu/drm/i915/intel_dsi_vbt.c +++ b/drivers/gpu/drm/i915/intel_dsi_vbt.c @@ -336,6 +336,12 @@ static void bxt_exec_gpio(struct drm_i915_private *dev_priv, gpiod_set_value(gpio_desc, value); } +static void icl_exec_gpio(struct drm_i915_private *dev_priv, + u8 gpio_source, u8 gpio_index, bool value) +{ + DRM_DEBUG_KMS("Skipping ICL GPIO element execution\n"); +} + static const u8 *mipi_exec_gpio(struct intel_dsi *intel_dsi, const u8 *data) { struct drm_device *dev = intel_dsi->base.base.dev; @@ -359,7 +365,9 @@ static const u8 *mipi_exec_gpio(struct intel_dsi *intel_dsi, const u8 *data) /* pull up/down */ value = *data++ & 1; - if (IS_VALLEYVIEW(dev_priv)) + if (IS_ICELAKE(dev_priv)) + icl_exec_gpio(dev_priv, gpio_source, gpio_index, value); + else if (IS_VALLEYVIEW(dev_priv)) vlv_exec_gpio(dev_priv, gpio_source, gpio_number, value); else if (IS_CHERRYVIEW(dev_priv)) chv_exec_gpio(dev_priv, gpio_source, gpio_number, value); From 32250c8e0ef909884b7d5fa4f6c7ea73a6125565 Mon Sep 17 00:00:00 2001 From: Madhav Chauhan Date: Thu, 29 Nov 2018 16:12:33 +0200 Subject: [PATCH 76/98] drm/i915/icl: Gate clocks for DSI As per BSPEC, depending on the DSI transcoder being used, DDI clock for the associated port should be gated. This patch does the same. Signed-off-by: Madhav Chauhan Signed-off-by: Vandita Kulkarni Signed-off-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/e032bc3d8fff91b8c2631c73121268214615a7e8.1543500286.git.jani.nikula@intel.com --- drivers/gpu/drm/i915/icl_dsi.c | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+) diff --git a/drivers/gpu/drm/i915/icl_dsi.c b/drivers/gpu/drm/i915/icl_dsi.c index 2b576a70222f..e42a99fda191 100644 --- a/drivers/gpu/drm/i915/icl_dsi.c +++ b/drivers/gpu/drm/i915/icl_dsi.c @@ -536,6 +536,23 @@ static void gen11_dsi_setup_dphy_timings(struct intel_encoder *encoder) } } +static void gen11_dsi_gate_clocks(struct intel_encoder *encoder) +{ + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); + struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base); + u32 tmp; + enum port port; + + mutex_lock(&dev_priv->dpll_lock); + tmp = I915_READ(DPCLKA_CFGCR0_ICL); + for_each_dsi_port(port, intel_dsi->ports) { + tmp |= DPCLKA_CFGCR0_DDI_CLK_OFF(port); + } + + I915_WRITE(DPCLKA_CFGCR0_ICL, tmp); + mutex_unlock(&dev_priv->dpll_lock); +} + static void gen11_dsi_configure_transcoder(struct intel_encoder *encoder, const struct intel_crtc_state *pipe_config) @@ -883,6 +900,9 @@ gen11_dsi_enable_port_and_phy(struct intel_encoder *encoder, /* Step (4h, 4i, 4j, 4k): Configure transcoder */ gen11_dsi_configure_transcoder(encoder, pipe_config); + + /* Step 4l: Gate DDI clocks */ + gen11_dsi_gate_clocks(encoder); } static void gen11_dsi_powerup_panel(struct intel_encoder *encoder) From 1026bea00381fe01326d220c55731ee150b89e10 Mon Sep 17 00:00:00 2001 From: Madhav Chauhan Date: Thu, 29 Nov 2018 16:12:34 +0200 Subject: [PATCH 77/98] drm/i915/icl: Ungate DSI clocks Ungate the clocks on the selected port. Signed-off-by: Madhav Chauhan Signed-off-by: Vandita Kulkarni Signed-off-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/2a01b3158feda624f98581b780854fe3df8c328d.1543500286.git.jani.nikula@intel.com --- drivers/gpu/drm/i915/icl_dsi.c | 19 +++++++++++++++++++ 1 file changed, 19 insertions(+) diff --git a/drivers/gpu/drm/i915/icl_dsi.c b/drivers/gpu/drm/i915/icl_dsi.c index e42a99fda191..e3aa9d3d2291 100644 --- a/drivers/gpu/drm/i915/icl_dsi.c +++ b/drivers/gpu/drm/i915/icl_dsi.c @@ -553,6 +553,23 @@ static void gen11_dsi_gate_clocks(struct intel_encoder *encoder) mutex_unlock(&dev_priv->dpll_lock); } +static void gen11_dsi_ungate_clocks(struct intel_encoder *encoder) +{ + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); + struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base); + u32 tmp; + enum port port; + + mutex_lock(&dev_priv->dpll_lock); + tmp = I915_READ(DPCLKA_CFGCR0_ICL); + for_each_dsi_port(port, intel_dsi->ports) { + tmp &= ~DPCLKA_CFGCR0_DDI_CLK_OFF(port); + } + + I915_WRITE(DPCLKA_CFGCR0_ICL, tmp); + mutex_unlock(&dev_priv->dpll_lock); +} + static void gen11_dsi_configure_transcoder(struct intel_encoder *encoder, const struct intel_crtc_state *pipe_config) @@ -1061,6 +1078,7 @@ static void gen11_dsi_disable_port(struct intel_encoder *encoder) u32 tmp; enum port port; + gen11_dsi_ungate_clocks(encoder); for_each_dsi_port(port, intel_dsi->ports) { tmp = I915_READ(DDI_BUF_CTL(port)); tmp &= ~DDI_BUF_CTL_ENABLE; @@ -1072,6 +1090,7 @@ static void gen11_dsi_disable_port(struct intel_encoder *encoder) DRM_ERROR("DDI port:%c buffer not idle\n", port_name(port)); } + gen11_dsi_ungate_clocks(encoder); } static void gen11_dsi_disable_io_power(struct intel_encoder *encoder) From 949fc52af19e5000073ee014b4f69ded137851ec Mon Sep 17 00:00:00 2001 From: Jani Nikula Date: Mon, 3 Dec 2018 11:43:26 +0200 Subject: [PATCH 78/98] drm/i915/icl: add pll mapping for DSI Add encoder specific pll mapping for DSI. The differences with the DDI version are big enough to warrant a separate function. v2: add posting read (Madhav) Cc: Madhav Chauhan Cc: Vandita Kulkarni Reviewed-by: Madhav Chauhan Signed-off-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/20181203094326.28294-1-jani.nikula@intel.com --- drivers/gpu/drm/i915/icl_dsi.c | 25 +++++++++++++++++++++++++ 1 file changed, 25 insertions(+) diff --git a/drivers/gpu/drm/i915/icl_dsi.c b/drivers/gpu/drm/i915/icl_dsi.c index e3aa9d3d2291..4dd793b78996 100644 --- a/drivers/gpu/drm/i915/icl_dsi.c +++ b/drivers/gpu/drm/i915/icl_dsi.c @@ -570,6 +570,28 @@ static void gen11_dsi_ungate_clocks(struct intel_encoder *encoder) mutex_unlock(&dev_priv->dpll_lock); } +static void gen11_dsi_map_pll(struct intel_encoder *encoder, + const struct intel_crtc_state *crtc_state) +{ + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); + struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base); + struct intel_shared_dpll *pll = crtc_state->shared_dpll; + enum port port; + u32 val; + + mutex_lock(&dev_priv->dpll_lock); + + val = I915_READ(DPCLKA_CFGCR0_ICL); + for_each_dsi_port(port, intel_dsi->ports) { + val &= ~DPCLKA_CFGCR0_DDI_CLK_SEL_MASK(port); + val |= DPCLKA_CFGCR0_DDI_CLK_SEL(pll->info->id, port); + } + I915_WRITE(DPCLKA_CFGCR0_ICL, val); + POSTING_READ(DPCLKA_CFGCR0_ICL); + + mutex_unlock(&dev_priv->dpll_lock); +} + static void gen11_dsi_configure_transcoder(struct intel_encoder *encoder, const struct intel_crtc_state *pipe_config) @@ -978,6 +1000,9 @@ static void gen11_dsi_pre_enable(struct intel_encoder *encoder, { struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base); + /* step3b */ + gen11_dsi_map_pll(encoder, pipe_config); + /* step4: enable DSI port and DPHY */ gen11_dsi_enable_port_and_phy(encoder, pipe_config); From 26af893184e58fb4898c09afbb548e0d893972bc Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Mon, 3 Dec 2018 11:36:54 +0000 Subject: [PATCH 79/98] drm/i915/breadcrumbs: Reduce missed-breadcrumb false positive rate Change the on-cpu check to on-runqueue to catch if the waiter has been woken (and reset its current_state back to TASK_UNINTERRUPTIBLE to perform the seqno check) but is sleeping due to being preempted off the cpu. Signed-off-by: Chris Wilson Cc: Tvrtko Ursulin Reviewed-by: Tvrtko Ursulin Link: https://patchwork.freedesktop.org/patch/msgid/20181203113701.12106-1-chris@chris-wilson.co.uk --- drivers/gpu/drm/i915/intel_breadcrumbs.c | 6 +----- 1 file changed, 1 insertion(+), 5 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_breadcrumbs.c b/drivers/gpu/drm/i915/intel_breadcrumbs.c index 84bf8d827136..447c5256f63a 100644 --- a/drivers/gpu/drm/i915/intel_breadcrumbs.c +++ b/drivers/gpu/drm/i915/intel_breadcrumbs.c @@ -27,11 +27,7 @@ #include "i915_drv.h" -#ifdef CONFIG_SMP -#define task_asleep(tsk) ((tsk)->state & TASK_NORMAL && !(tsk)->on_cpu) -#else -#define task_asleep(tsk) ((tsk)->state & TASK_NORMAL) -#endif +#define task_asleep(tsk) ((tsk)->state & TASK_NORMAL && !(tsk)->on_rq) static unsigned int __intel_breadcrumbs_wakeup(struct intel_breadcrumbs *b) { From 46592892e1a60f9e9de3287719143a148fce93cf Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Fri, 30 Nov 2018 12:59:54 +0000 Subject: [PATCH 80/98] drm/i915/vgpu: Disallow loading on old vGPU hosts Since commit fd8526e50902 ("drm/i915/execlists: Trust the CSB") we actually broke the force-mmio mode for our execlists implementation. No one noticed, so ergo no one is actually using an old vGPU host (where we required the older method) and so can simply remove the broken support. v2: csb_read can go as well (Mika) Reported-by: Mika Kuoppala Fixes: fd8526e50902 ("drm/i915/execlists: Trust the CSB") Signed-off-by: Chris Wilson Cc: Tvrtko Ursulin Cc: Joonas Lahtinen Cc: Mika Kuoppala Cc: Daniele Ceraolo Spurio Reviewed-by: Mika Kuoppala Link: https://patchwork.freedesktop.org/patch/msgid/20181130125954.11924-1-chris@chris-wilson.co.uk --- drivers/gpu/drm/i915/i915_drv.c | 14 +++++++++++ drivers/gpu/drm/i915/intel_lrc.c | 32 +++++++------------------ drivers/gpu/drm/i915/intel_ringbuffer.h | 16 ------------- 3 files changed, 22 insertions(+), 40 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c index e39016713464..3e5e2efce670 100644 --- a/drivers/gpu/drm/i915/i915_drv.c +++ b/drivers/gpu/drm/i915/i915_drv.c @@ -1384,6 +1384,20 @@ static int i915_driver_init_hw(struct drm_i915_private *dev_priv) } } + if (HAS_EXECLISTS(dev_priv)) { + /* + * Older GVT emulation depends upon intercepting CSB mmio, + * which we no longer use, preferring to use the HWSP cache + * instead. + */ + if (intel_vgpu_active(dev_priv) && + !intel_vgpu_has_hwsp_emulation(dev_priv)) { + i915_report_error(dev_priv, + "old vGPU host found, support for HWSP emulation required\n"); + return -ENXIO; + } + } + intel_sanitize_options(dev_priv); i915_perf_init(dev_priv); diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c index 11f4e6148557..1f004683b777 100644 --- a/drivers/gpu/drm/i915/intel_lrc.c +++ b/drivers/gpu/drm/i915/intel_lrc.c @@ -767,6 +767,8 @@ execlists_cancel_port_requests(struct intel_engine_execlists * const execlists) static void reset_csb_pointers(struct intel_engine_execlists *execlists) { + const unsigned int reset_value = GEN8_CSB_ENTRIES - 1; + /* * After a reset, the HW starts writing into CSB entry [0]. We * therefore have to set our HEAD pointer back one entry so that @@ -776,8 +778,8 @@ static void reset_csb_pointers(struct intel_engine_execlists *execlists) * inline comparison of our cached head position against the last HW * write works even before the first interrupt. */ - execlists->csb_head = execlists->csb_write_reset; - WRITE_ONCE(*execlists->csb_write, execlists->csb_write_reset); + execlists->csb_head = reset_value; + WRITE_ONCE(*execlists->csb_write, reset_value); } static void nop_submission_tasklet(unsigned long data) @@ -2217,12 +2219,6 @@ logical_ring_setup(struct intel_engine_cs *engine) logical_ring_default_irqs(engine); } -static bool csb_force_mmio(struct drm_i915_private *i915) -{ - /* Older GVT emulation depends upon intercepting CSB mmio */ - return intel_vgpu_active(i915) && !intel_vgpu_has_hwsp_emulation(i915); -} - static int logical_ring_init(struct intel_engine_cs *engine) { struct drm_i915_private *i915 = engine->i915; @@ -2252,24 +2248,12 @@ static int logical_ring_init(struct intel_engine_cs *engine) upper_32_bits(ce->lrc_desc); } - execlists->csb_read = - i915->regs + i915_mmio_reg_offset(RING_CONTEXT_STATUS_PTR(engine)); - if (csb_force_mmio(i915)) { - execlists->csb_status = (u32 __force *) - (i915->regs + i915_mmio_reg_offset(RING_CONTEXT_STATUS_BUF_LO(engine, 0))); + execlists->csb_status = + &engine->status_page.page_addr[I915_HWS_CSB_BUF0_INDEX]; - execlists->csb_write = (u32 __force *)execlists->csb_read; - execlists->csb_write_reset = - _MASKED_FIELD(GEN8_CSB_WRITE_PTR_MASK, - GEN8_CSB_ENTRIES - 1); - } else { - execlists->csb_status = - &engine->status_page.page_addr[I915_HWS_CSB_BUF0_INDEX]; + execlists->csb_write = + &engine->status_page.page_addr[intel_hws_csb_write_index(i915)]; - execlists->csb_write = - &engine->status_page.page_addr[intel_hws_csb_write_index(i915)]; - execlists->csb_write_reset = GEN8_CSB_ENTRIES - 1; - } reset_csb_pointers(execlists); return 0; diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h index 8a2270b209b0..f4988f7bc756 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.h +++ b/drivers/gpu/drm/i915/intel_ringbuffer.h @@ -312,13 +312,6 @@ struct intel_engine_execlists { */ struct rb_root_cached queue; - /** - * @csb_read: control register for Context Switch buffer - * - * Note this register is always in mmio. - */ - u32 __iomem *csb_read; - /** * @csb_write: control register for Context Switch buffer * @@ -338,15 +331,6 @@ struct intel_engine_execlists { */ u32 preempt_complete_status; - /** - * @csb_write_reset: reset value for CSB write pointer - * - * As the CSB write pointer maybe either in HWSP or as a field - * inside an mmio register, we want to reprogram it slightly - * differently to avoid later confusion. - */ - u32 csb_write_reset; - /** * @csb_head: context status buffer head */ From 9ee4685c9ac591b71af755657c3f6ce428ebcca4 Mon Sep 17 00:00:00 2001 From: Jani Nikula Date: Thu, 4 Oct 2018 17:37:49 +0300 Subject: [PATCH 81/98] sysfs: constify sysfs create/remove files harder Let the passed in array be const (and thus placed in rodata) instead of a mutable array of const pointers. Cc: Greg Kroah-Hartman Cc: "Rafael J. Wysocki" Reviewed-by: Rafael J. Wysocki Reviewed-by: Greg Kroah-Hartman Signed-off-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/20181004143750.30880-1-jani.nikula@intel.com --- fs/sysfs/file.c | 4 ++-- include/linux/sysfs.h | 8 ++++---- 2 files changed, 6 insertions(+), 6 deletions(-) diff --git a/fs/sysfs/file.c b/fs/sysfs/file.c index 0a7252aecfa5..bb71db63c99c 100644 --- a/fs/sysfs/file.c +++ b/fs/sysfs/file.c @@ -334,7 +334,7 @@ int sysfs_create_file_ns(struct kobject *kobj, const struct attribute *attr, } EXPORT_SYMBOL_GPL(sysfs_create_file_ns); -int sysfs_create_files(struct kobject *kobj, const struct attribute **ptr) +int sysfs_create_files(struct kobject *kobj, const struct attribute * const *ptr) { int err = 0; int i; @@ -493,7 +493,7 @@ bool sysfs_remove_file_self(struct kobject *kobj, const struct attribute *attr) return ret; } -void sysfs_remove_files(struct kobject *kobj, const struct attribute **ptr) +void sysfs_remove_files(struct kobject *kobj, const struct attribute * const *ptr) { int i; for (i = 0; ptr[i]; i++) diff --git a/include/linux/sysfs.h b/include/linux/sysfs.h index 987cefa337de..786816cf4aa5 100644 --- a/include/linux/sysfs.h +++ b/include/linux/sysfs.h @@ -234,7 +234,7 @@ int __must_check sysfs_create_file_ns(struct kobject *kobj, const struct attribute *attr, const void *ns); int __must_check sysfs_create_files(struct kobject *kobj, - const struct attribute **attr); + const struct attribute * const *attr); int __must_check sysfs_chmod_file(struct kobject *kobj, const struct attribute *attr, umode_t mode); struct kernfs_node *sysfs_break_active_protection(struct kobject *kobj, @@ -243,7 +243,7 @@ void sysfs_unbreak_active_protection(struct kernfs_node *kn); void sysfs_remove_file_ns(struct kobject *kobj, const struct attribute *attr, const void *ns); bool sysfs_remove_file_self(struct kobject *kobj, const struct attribute *attr); -void sysfs_remove_files(struct kobject *kobj, const struct attribute **attr); +void sysfs_remove_files(struct kobject *kobj, const struct attribute * const *attr); int __must_check sysfs_create_bin_file(struct kobject *kobj, const struct bin_attribute *attr); @@ -342,7 +342,7 @@ static inline int sysfs_create_file_ns(struct kobject *kobj, } static inline int sysfs_create_files(struct kobject *kobj, - const struct attribute **attr) + const struct attribute * const *attr) { return 0; } @@ -377,7 +377,7 @@ static inline bool sysfs_remove_file_self(struct kobject *kobj, } static inline void sysfs_remove_files(struct kobject *kobj, - const struct attribute **attr) + const struct attribute * const *attr) { } From e1215de823ca50ba1ca313c35ec75abad0362593 Mon Sep 17 00:00:00 2001 From: Jani Nikula Date: Thu, 4 Oct 2018 17:37:50 +0300 Subject: [PATCH 82/98] drm/i915/sysfs: make attrs arrays const They don't need to be modified. Cc: Greg Kroah-Hartman Cc: "Rafael J. Wysocki" Reviewed-by: Rafael J. Wysocki Signed-off-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/20181004143750.30880-2-jani.nikula@intel.com --- drivers/gpu/drm/i915/i915_sysfs.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_sysfs.c b/drivers/gpu/drm/i915/i915_sysfs.c index ae63a7d0f51d..535caebd9813 100644 --- a/drivers/gpu/drm/i915/i915_sysfs.c +++ b/drivers/gpu/drm/i915/i915_sysfs.c @@ -483,7 +483,7 @@ static ssize_t gt_rp_mhz_show(struct device *kdev, struct device_attribute *attr return snprintf(buf, PAGE_SIZE, "%d\n", val); } -static const struct attribute *gen6_attrs[] = { +static const struct attribute * const gen6_attrs[] = { &dev_attr_gt_act_freq_mhz.attr, &dev_attr_gt_cur_freq_mhz.attr, &dev_attr_gt_boost_freq_mhz.attr, @@ -495,7 +495,7 @@ static const struct attribute *gen6_attrs[] = { NULL, }; -static const struct attribute *vlv_attrs[] = { +static const struct attribute * const vlv_attrs[] = { &dev_attr_gt_act_freq_mhz.attr, &dev_attr_gt_cur_freq_mhz.attr, &dev_attr_gt_boost_freq_mhz.attr, From ae9e7ced4f7b0e9d1646844b4713d6b60e9b3f65 Mon Sep 17 00:00:00 2001 From: Manasi Navare Date: Fri, 30 Nov 2018 17:04:12 -0800 Subject: [PATCH 83/98] drm/i915/dp: Fix link compute m_n calc for DSC MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fix the intel_link_compute_m_n in case of display stream compression. This patch passes the compressed_bpp to intel_link_compute_m_n if compression is enabled. Fixes: a4a157777c80 ("drm/i915/dp: Compute DSC pipe config in atomic check") Cc: Ville Syrjala Cc: Anusha Srivatsa Signed-off-by: Manasi Navare Reviewed-by: Ville Syrjälä Link: https://patchwork.freedesktop.org/patch/msgid/20181201010412.32372-1-manasi.d.navare@intel.com --- drivers/gpu/drm/i915/intel_dp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 38a6e82153fd..a6907a1761ab 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -2173,7 +2173,7 @@ intel_dp_compute_config(struct intel_encoder *encoder, &pipe_config->dp_m_n, constant_n); else - intel_link_compute_m_n(pipe_config->dsc_params.compression_enable, + intel_link_compute_m_n(pipe_config->dsc_params.compressed_bpp, pipe_config->lane_count, adjusted_mode->crtc_clock, pipe_config->port_clock, From e1bf094b3c7514934168dcb69f18b65aacb5a953 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Jos=C3=A9=20Roberto=20de=20Souza?= Date: Fri, 30 Nov 2018 15:20:47 -0800 Subject: [PATCH 84/98] drm/i915: Add HAS_DISPLAY() and use it MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Right now it is decided if GEN has display by checking the num_pipes, so lets make it explicit and use a macro. Cc: Jani Nikula Reviewed-by: Lucas De Marchi Signed-off-by: José Roberto de Souza Link: https://patchwork.freedesktop.org/patch/msgid/20181130232048.14216-1-jose.souza@intel.com --- drivers/gpu/drm/i915/i915_drv.c | 10 +++++----- drivers/gpu/drm/i915/i915_drv.h | 2 ++ drivers/gpu/drm/i915/intel_bios.c | 2 +- drivers/gpu/drm/i915/intel_device_info.c | 4 ++-- drivers/gpu/drm/i915/intel_display.c | 4 ++-- drivers/gpu/drm/i915/intel_fbdev.c | 2 +- drivers/gpu/drm/i915/intel_i2c.c | 2 +- 7 files changed, 14 insertions(+), 12 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c index 3e5e2efce670..c2a3ba7e09e0 100644 --- a/drivers/gpu/drm/i915/i915_drv.c +++ b/drivers/gpu/drm/i915/i915_drv.c @@ -287,7 +287,7 @@ static void intel_detect_pch(struct drm_i915_private *dev_priv) * Use PCH_NOP (PCH but no South Display) for PCH platforms without * display. */ - if (pch && INTEL_INFO(dev_priv)->num_pipes == 0) { + if (pch && !HAS_DISPLAY(dev_priv)) { DRM_DEBUG_KMS("Display disabled, reverting to NOP PCH\n"); dev_priv->pch_type = PCH_NOP; dev_priv->pch_id = 0; @@ -645,7 +645,7 @@ static int i915_load_modeset_init(struct drm_device *dev) if (i915_inject_load_failure()) return -ENODEV; - if (INTEL_INFO(dev_priv)->num_pipes) { + if (HAS_DISPLAY(dev_priv)) { ret = drm_vblank_init(&dev_priv->drm, INTEL_INFO(dev_priv)->num_pipes); if (ret) @@ -696,7 +696,7 @@ static int i915_load_modeset_init(struct drm_device *dev) intel_overlay_setup(dev_priv); - if (INTEL_INFO(dev_priv)->num_pipes == 0) + if (!HAS_DISPLAY(dev_priv)) return 0; ret = intel_fbdev_init(dev); @@ -1566,7 +1566,7 @@ static void i915_driver_register(struct drm_i915_private *dev_priv) } else DRM_ERROR("Failed to register driver for userspace access!\n"); - if (INTEL_INFO(dev_priv)->num_pipes) { + if (HAS_DISPLAY(dev_priv)) { /* Must be done after probing outputs */ intel_opregion_register(dev_priv); acpi_video_register(); @@ -1590,7 +1590,7 @@ static void i915_driver_register(struct drm_i915_private *dev_priv) * We need to coordinate the hotplugs with the asynchronous fbdev * configuration, for which we use the fbdev->async_cookie. */ - if (INTEL_INFO(dev_priv)->num_pipes) + if (HAS_DISPLAY(dev_priv)) drm_kms_helper_poll_init(dev); intel_power_domains_enable(dev_priv); diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index d45475287130..f38658e1a20a 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -2569,6 +2569,8 @@ intel_info(const struct drm_i915_private *dev_priv) #define GT_FREQUENCY_MULTIPLIER 50 #define GEN9_FREQ_SCALER 3 +#define HAS_DISPLAY(dev_priv) (INTEL_INFO(dev_priv)->num_pipes > 0) + #include "i915_trace.h" static inline bool intel_vtd_active(void) diff --git a/drivers/gpu/drm/i915/intel_bios.c b/drivers/gpu/drm/i915/intel_bios.c index 0694aa8bb9bc..6d3e0260d49c 100644 --- a/drivers/gpu/drm/i915/intel_bios.c +++ b/drivers/gpu/drm/i915/intel_bios.c @@ -1752,7 +1752,7 @@ void intel_bios_init(struct drm_i915_private *dev_priv) const struct bdb_header *bdb; u8 __iomem *bios = NULL; - if (INTEL_INFO(dev_priv)->num_pipes == 0) { + if (!HAS_DISPLAY(dev_priv)) { DRM_DEBUG_KMS("Skipping VBT init due to disabled display.\n"); return; } diff --git a/drivers/gpu/drm/i915/intel_device_info.c b/drivers/gpu/drm/i915/intel_device_info.c index ceecb5bd5226..677002a9e893 100644 --- a/drivers/gpu/drm/i915/intel_device_info.c +++ b/drivers/gpu/drm/i915/intel_device_info.c @@ -782,7 +782,7 @@ void intel_device_info_runtime_init(struct intel_device_info *info) if (i915_modparams.disable_display) { DRM_INFO("Display disabled (module parameter)\n"); info->num_pipes = 0; - } else if (info->num_pipes > 0 && + } else if (HAS_DISPLAY(dev_priv) && (IS_GEN7(dev_priv) || IS_GEN8(dev_priv)) && HAS_PCH_SPLIT(dev_priv)) { u32 fuse_strap = I915_READ(FUSE_STRAP); @@ -807,7 +807,7 @@ void intel_device_info_runtime_init(struct intel_device_info *info) DRM_INFO("PipeC fused off\n"); info->num_pipes -= 1; } - } else if (info->num_pipes > 0 && IS_GEN9(dev_priv)) { + } else if (HAS_DISPLAY(dev_priv) && IS_GEN9(dev_priv)) { u32 dfsm = I915_READ(SKL_DFSM); u8 disabled_mask = 0; bool invalid; diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 58cadd251acc..aa5f89f26cf5 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -14249,7 +14249,7 @@ static void intel_setup_outputs(struct drm_i915_private *dev_priv) intel_pps_init(dev_priv); - if (INTEL_INFO(dev_priv)->num_pipes == 0) + if (!HAS_DISPLAY(dev_priv)) return; /* @@ -16040,7 +16040,7 @@ intel_display_capture_error_state(struct drm_i915_private *dev_priv) }; int i; - if (INTEL_INFO(dev_priv)->num_pipes == 0) + if (!HAS_DISPLAY(dev_priv)) return NULL; error = kzalloc(sizeof(*error), GFP_ATOMIC); diff --git a/drivers/gpu/drm/i915/intel_fbdev.c b/drivers/gpu/drm/i915/intel_fbdev.c index 2480c7d6edee..fb5bb5b32a60 100644 --- a/drivers/gpu/drm/i915/intel_fbdev.c +++ b/drivers/gpu/drm/i915/intel_fbdev.c @@ -672,7 +672,7 @@ int intel_fbdev_init(struct drm_device *dev) struct intel_fbdev *ifbdev; int ret; - if (WARN_ON(INTEL_INFO(dev_priv)->num_pipes == 0)) + if (WARN_ON(!HAS_DISPLAY(dev_priv))) return -ENODEV; ifbdev = kzalloc(sizeof(struct intel_fbdev), GFP_KERNEL); diff --git a/drivers/gpu/drm/i915/intel_i2c.c b/drivers/gpu/drm/i915/intel_i2c.c index 33d87ab93fdd..802d0394ccc4 100644 --- a/drivers/gpu/drm/i915/intel_i2c.c +++ b/drivers/gpu/drm/i915/intel_i2c.c @@ -817,7 +817,7 @@ int intel_setup_gmbus(struct drm_i915_private *dev_priv) unsigned int pin; int ret; - if (INTEL_INFO(dev_priv)->num_pipes == 0) + if (!HAS_DISPLAY(dev_priv)) return 0; if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) From d53db442db36fdba1a6584bd5cf49124ba17d6e9 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Jos=C3=A9=20Roberto=20de=20Souza?= Date: Fri, 30 Nov 2018 15:20:48 -0800 Subject: [PATCH 85/98] drm/i915: Move display device info capabilities to its own struct MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This helps separate what capabilities are display capabilities. v3: Moving display struct right after flags (Lucas) Cc: Jani Nikula Suggested-by: Jani Nikula Suggested-by: Lucas De Marchi Reviewed-by: Lucas De Marchi Signed-off-by: José Roberto de Souza Link: https://patchwork.freedesktop.org/patch/msgid/20181130232048.14216-2-jose.souza@intel.com --- drivers/gpu/drm/i915/i915_drv.h | 22 ++--- drivers/gpu/drm/i915/i915_pci.c | 117 +++++++++++++---------- drivers/gpu/drm/i915/intel_device_info.c | 4 + drivers/gpu/drm/i915/intel_device_info.h | 32 ++++--- drivers/gpu/drm/i915/intel_display.c | 4 +- drivers/gpu/drm/i915/intel_fbc.c | 2 +- 6 files changed, 103 insertions(+), 78 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index f38658e1a20a..12b8476b09a2 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -2449,9 +2449,9 @@ intel_info(const struct drm_i915_private *dev_priv) ((sizes) & ~(dev_priv)->info.page_sizes) == 0; \ }) -#define HAS_OVERLAY(dev_priv) ((dev_priv)->info.has_overlay) +#define HAS_OVERLAY(dev_priv) ((dev_priv)->info.display.has_overlay) #define OVERLAY_NEEDS_PHYSICAL(dev_priv) \ - ((dev_priv)->info.overlay_needs_physical) + ((dev_priv)->info.display.overlay_needs_physical) /* Early gen2 have a totally busted CS tlb and require pinned batches. */ #define HAS_BROKEN_CS_TLB(dev_priv) (IS_I830(dev_priv) || IS_I845G(dev_priv)) @@ -2472,31 +2472,31 @@ intel_info(const struct drm_i915_private *dev_priv) #define HAS_128_BYTE_Y_TILING(dev_priv) (!IS_GEN2(dev_priv) && \ !(IS_I915G(dev_priv) || \ IS_I915GM(dev_priv))) -#define SUPPORTS_TV(dev_priv) ((dev_priv)->info.supports_tv) -#define I915_HAS_HOTPLUG(dev_priv) ((dev_priv)->info.has_hotplug) +#define SUPPORTS_TV(dev_priv) ((dev_priv)->info.display.supports_tv) +#define I915_HAS_HOTPLUG(dev_priv) ((dev_priv)->info.display.has_hotplug) #define HAS_FW_BLC(dev_priv) (INTEL_GEN(dev_priv) > 2) -#define HAS_FBC(dev_priv) ((dev_priv)->info.has_fbc) +#define HAS_FBC(dev_priv) ((dev_priv)->info.display.has_fbc) #define HAS_CUR_FBC(dev_priv) (!HAS_GMCH_DISPLAY(dev_priv) && INTEL_GEN(dev_priv) >= 7) #define HAS_IPS(dev_priv) (IS_HSW_ULT(dev_priv) || IS_BROADWELL(dev_priv)) -#define HAS_DP_MST(dev_priv) ((dev_priv)->info.has_dp_mst) +#define HAS_DP_MST(dev_priv) ((dev_priv)->info.display.has_dp_mst) -#define HAS_DDI(dev_priv) ((dev_priv)->info.has_ddi) +#define HAS_DDI(dev_priv) ((dev_priv)->info.display.has_ddi) #define HAS_FPGA_DBG_UNCLAIMED(dev_priv) ((dev_priv)->info.has_fpga_dbg) -#define HAS_PSR(dev_priv) ((dev_priv)->info.has_psr) +#define HAS_PSR(dev_priv) ((dev_priv)->info.display.has_psr) #define HAS_RC6(dev_priv) ((dev_priv)->info.has_rc6) #define HAS_RC6p(dev_priv) ((dev_priv)->info.has_rc6p) #define HAS_RC6pp(dev_priv) (false) /* HW was never validated */ -#define HAS_CSR(dev_priv) ((dev_priv)->info.has_csr) +#define HAS_CSR(dev_priv) ((dev_priv)->info.display.has_csr) #define HAS_RUNTIME_PM(dev_priv) ((dev_priv)->info.has_runtime_pm) #define HAS_64BIT_RELOC(dev_priv) ((dev_priv)->info.has_64bit_reloc) -#define HAS_IPC(dev_priv) ((dev_priv)->info.has_ipc) +#define HAS_IPC(dev_priv) ((dev_priv)->info.display.has_ipc) /* * For now, anything with a GuC requires uCode loading, and then supports @@ -2557,7 +2557,7 @@ intel_info(const struct drm_i915_private *dev_priv) #define HAS_PCH_NOP(dev_priv) (INTEL_PCH_TYPE(dev_priv) == PCH_NOP) #define HAS_PCH_SPLIT(dev_priv) (INTEL_PCH_TYPE(dev_priv) != PCH_NONE) -#define HAS_GMCH_DISPLAY(dev_priv) ((dev_priv)->info.has_gmch_display) +#define HAS_GMCH_DISPLAY(dev_priv) ((dev_priv)->info.display.has_gmch_display) #define HAS_LSPCON(dev_priv) (INTEL_GEN(dev_priv) >= 9) diff --git a/drivers/gpu/drm/i915/i915_pci.c b/drivers/gpu/drm/i915/i915_pci.c index 1b81d7cb209e..6350db5503cd 100644 --- a/drivers/gpu/drm/i915/i915_pci.c +++ b/drivers/gpu/drm/i915/i915_pci.c @@ -79,8 +79,9 @@ #define GEN2_FEATURES \ GEN(2), \ .num_pipes = 1, \ - .has_overlay = 1, .overlay_needs_physical = 1, \ - .has_gmch_display = 1, \ + .display.has_overlay = 1, \ + .display.overlay_needs_physical = 1, \ + .display.has_gmch_display = 1, \ .hws_needs_physical = 1, \ .unfenced_needs_alignment = 1, \ .ring_mask = RENDER_RING, \ @@ -93,7 +94,8 @@ static const struct intel_device_info intel_i830_info = { GEN2_FEATURES, PLATFORM(INTEL_I830), - .is_mobile = 1, .cursor_needs_physical = 1, + .is_mobile = 1, + .display.cursor_needs_physical = 1, .num_pipes = 2, /* legal, last one wins */ }; @@ -107,8 +109,8 @@ static const struct intel_device_info intel_i85x_info = { PLATFORM(INTEL_I85X), .is_mobile = 1, .num_pipes = 2, /* legal, last one wins */ - .cursor_needs_physical = 1, - .has_fbc = 1, + .display.cursor_needs_physical = 1, + .display.has_fbc = 1, }; static const struct intel_device_info intel_i865g_info = { @@ -119,7 +121,7 @@ static const struct intel_device_info intel_i865g_info = { #define GEN3_FEATURES \ GEN(3), \ .num_pipes = 2, \ - .has_gmch_display = 1, \ + .display.has_gmch_display = 1, \ .ring_mask = RENDER_RING, \ .has_snoop = true, \ .has_coherent_ggtt = true, \ @@ -131,8 +133,9 @@ static const struct intel_device_info intel_i915g_info = { GEN3_FEATURES, PLATFORM(INTEL_I915G), .has_coherent_ggtt = false, - .cursor_needs_physical = 1, - .has_overlay = 1, .overlay_needs_physical = 1, + .display.cursor_needs_physical = 1, + .display.has_overlay = 1, + .display.overlay_needs_physical = 1, .hws_needs_physical = 1, .unfenced_needs_alignment = 1, }; @@ -141,10 +144,11 @@ static const struct intel_device_info intel_i915gm_info = { GEN3_FEATURES, PLATFORM(INTEL_I915GM), .is_mobile = 1, - .cursor_needs_physical = 1, - .has_overlay = 1, .overlay_needs_physical = 1, - .supports_tv = 1, - .has_fbc = 1, + .display.cursor_needs_physical = 1, + .display.has_overlay = 1, + .display.overlay_needs_physical = 1, + .display.supports_tv = 1, + .display.has_fbc = 1, .hws_needs_physical = 1, .unfenced_needs_alignment = 1, }; @@ -152,8 +156,10 @@ static const struct intel_device_info intel_i915gm_info = { static const struct intel_device_info intel_i945g_info = { GEN3_FEATURES, PLATFORM(INTEL_I945G), - .has_hotplug = 1, .cursor_needs_physical = 1, - .has_overlay = 1, .overlay_needs_physical = 1, + .display.has_hotplug = 1, + .display.cursor_needs_physical = 1, + .display.has_overlay = 1, + .display.overlay_needs_physical = 1, .hws_needs_physical = 1, .unfenced_needs_alignment = 1, }; @@ -162,10 +168,12 @@ static const struct intel_device_info intel_i945gm_info = { GEN3_FEATURES, PLATFORM(INTEL_I945GM), .is_mobile = 1, - .has_hotplug = 1, .cursor_needs_physical = 1, - .has_overlay = 1, .overlay_needs_physical = 1, - .supports_tv = 1, - .has_fbc = 1, + .display.has_hotplug = 1, + .display.cursor_needs_physical = 1, + .display.has_overlay = 1, + .display.overlay_needs_physical = 1, + .display.supports_tv = 1, + .display.has_fbc = 1, .hws_needs_physical = 1, .unfenced_needs_alignment = 1, }; @@ -173,23 +181,23 @@ static const struct intel_device_info intel_i945gm_info = { static const struct intel_device_info intel_g33_info = { GEN3_FEATURES, PLATFORM(INTEL_G33), - .has_hotplug = 1, - .has_overlay = 1, + .display.has_hotplug = 1, + .display.has_overlay = 1, }; static const struct intel_device_info intel_pineview_info = { GEN3_FEATURES, PLATFORM(INTEL_PINEVIEW), .is_mobile = 1, - .has_hotplug = 1, - .has_overlay = 1, + .display.has_hotplug = 1, + .display.has_overlay = 1, }; #define GEN4_FEATURES \ GEN(4), \ .num_pipes = 2, \ - .has_hotplug = 1, \ - .has_gmch_display = 1, \ + .display.has_hotplug = 1, \ + .display.has_gmch_display = 1, \ .ring_mask = RENDER_RING, \ .has_snoop = true, \ .has_coherent_ggtt = true, \ @@ -200,7 +208,7 @@ static const struct intel_device_info intel_pineview_info = { static const struct intel_device_info intel_i965g_info = { GEN4_FEATURES, PLATFORM(INTEL_I965G), - .has_overlay = 1, + .display.has_overlay = 1, .hws_needs_physical = 1, .has_snoop = false, }; @@ -208,9 +216,10 @@ static const struct intel_device_info intel_i965g_info = { static const struct intel_device_info intel_i965gm_info = { GEN4_FEATURES, PLATFORM(INTEL_I965GM), - .is_mobile = 1, .has_fbc = 1, - .has_overlay = 1, - .supports_tv = 1, + .is_mobile = 1, + .display.has_fbc = 1, + .display.has_overlay = 1, + .display.supports_tv = 1, .hws_needs_physical = 1, .has_snoop = false, }; @@ -224,15 +233,16 @@ static const struct intel_device_info intel_g45_info = { static const struct intel_device_info intel_gm45_info = { GEN4_FEATURES, PLATFORM(INTEL_GM45), - .is_mobile = 1, .has_fbc = 1, - .supports_tv = 1, + .is_mobile = 1, + .display.has_fbc = 1, + .display.supports_tv = 1, .ring_mask = RENDER_RING | BSD_RING, }; #define GEN5_FEATURES \ GEN(5), \ .num_pipes = 2, \ - .has_hotplug = 1, \ + .display.has_hotplug = 1, \ .ring_mask = RENDER_RING | BSD_RING, \ .has_snoop = true, \ .has_coherent_ggtt = true, \ @@ -250,14 +260,15 @@ static const struct intel_device_info intel_ironlake_d_info = { static const struct intel_device_info intel_ironlake_m_info = { GEN5_FEATURES, PLATFORM(INTEL_IRONLAKE), - .is_mobile = 1, .has_fbc = 1, + .is_mobile = 1, + .display.has_fbc = 1, }; #define GEN6_FEATURES \ GEN(6), \ .num_pipes = 2, \ - .has_hotplug = 1, \ - .has_fbc = 1, \ + .display.has_hotplug = 1, \ + .display.has_fbc = 1, \ .ring_mask = RENDER_RING | BSD_RING | BLT_RING, \ .has_coherent_ggtt = true, \ .has_llc = 1, \ @@ -301,8 +312,8 @@ static const struct intel_device_info intel_sandybridge_m_gt2_info = { #define GEN7_FEATURES \ GEN(7), \ .num_pipes = 3, \ - .has_hotplug = 1, \ - .has_fbc = 1, \ + .display.has_hotplug = 1, \ + .display.has_fbc = 1, \ .ring_mask = RENDER_RING | BSD_RING | BLT_RING, \ .has_coherent_ggtt = true, \ .has_llc = 1, \ @@ -359,8 +370,8 @@ static const struct intel_device_info intel_valleyview_info = { .num_pipes = 2, .has_runtime_pm = 1, .has_rc6 = 1, - .has_gmch_display = 1, - .has_hotplug = 1, + .display.has_gmch_display = 1, + .display.has_hotplug = 1, .ppgtt = INTEL_PPGTT_FULL, .has_snoop = true, .has_coherent_ggtt = false, @@ -374,10 +385,10 @@ static const struct intel_device_info intel_valleyview_info = { #define G75_FEATURES \ GEN7_FEATURES, \ .ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING, \ - .has_ddi = 1, \ + .display.has_ddi = 1, \ .has_fpga_dbg = 1, \ - .has_psr = 1, \ - .has_dp_mst = 1, \ + .display.has_psr = 1, \ + .display.has_dp_mst = 1, \ .has_rc6p = 0 /* RC6p removed-by HSW */, \ .has_runtime_pm = 1 @@ -444,14 +455,14 @@ static const struct intel_device_info intel_cherryview_info = { PLATFORM(INTEL_CHERRYVIEW), GEN(8), .num_pipes = 3, - .has_hotplug = 1, + .display.has_hotplug = 1, .is_lp = 1, .ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING, .has_64bit_reloc = 1, .has_runtime_pm = 1, .has_rc6 = 1, .has_logical_ring_contexts = 1, - .has_gmch_display = 1, + .display.has_gmch_display = 1, .ppgtt = INTEL_PPGTT_FULL, .has_reset_engine = 1, .has_snoop = true, @@ -473,15 +484,15 @@ static const struct intel_device_info intel_cherryview_info = { GEN(9), \ GEN9_DEFAULT_PAGE_SIZES, \ .has_logical_ring_preemption = 1, \ - .has_csr = 1, \ + .display.has_csr = 1, \ .has_guc = 1, \ - .has_ipc = 1, \ + .display.has_ipc = 1, \ .ddb_size = 896 #define SKL_PLATFORM \ GEN9_FEATURES, \ /* Display WA #0477 WaDisableIPC: skl */ \ - .has_ipc = 0, \ + .display.has_ipc = 0, \ PLATFORM(INTEL_SKYLAKE) static const struct intel_device_info intel_skylake_gt1_info = { @@ -512,19 +523,19 @@ static const struct intel_device_info intel_skylake_gt4_info = { #define GEN9_LP_FEATURES \ GEN(9), \ .is_lp = 1, \ - .has_hotplug = 1, \ + .display.has_hotplug = 1, \ .ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING, \ .num_pipes = 3, \ .has_64bit_reloc = 1, \ - .has_ddi = 1, \ + .display.has_ddi = 1, \ .has_fpga_dbg = 1, \ - .has_fbc = 1, \ - .has_psr = 1, \ + .display.has_fbc = 1, \ + .display.has_psr = 1, \ .has_runtime_pm = 1, \ .has_pooled_eu = 0, \ - .has_csr = 1, \ + .display.has_csr = 1, \ .has_rc6 = 1, \ - .has_dp_mst = 1, \ + .display.has_dp_mst = 1, \ .has_logical_ring_contexts = 1, \ .has_logical_ring_preemption = 1, \ .has_guc = 1, \ @@ -532,7 +543,7 @@ static const struct intel_device_info intel_skylake_gt4_info = { .has_reset_engine = 1, \ .has_snoop = true, \ .has_coherent_ggtt = false, \ - .has_ipc = 1, \ + .display.has_ipc = 1, \ GEN9_DEFAULT_PAGE_SIZES, \ GEN_DEFAULT_PIPEOFFSETS, \ IVB_CURSOR_OFFSETS, \ diff --git a/drivers/gpu/drm/i915/intel_device_info.c b/drivers/gpu/drm/i915/intel_device_info.c index 677002a9e893..1e56319334f3 100644 --- a/drivers/gpu/drm/i915/intel_device_info.c +++ b/drivers/gpu/drm/i915/intel_device_info.c @@ -77,6 +77,10 @@ void intel_device_info_dump_flags(const struct intel_device_info *info, #define PRINT_FLAG(name) drm_printf(p, "%s: %s\n", #name, yesno(info->name)); DEV_INFO_FOR_EACH_FLAG(PRINT_FLAG); #undef PRINT_FLAG + +#define PRINT_FLAG(name) drm_printf(p, "%s: %s\n", #name, yesno(info->display.name)); + DEV_INFO_DISPLAY_FOR_EACH_FLAG(PRINT_FLAG); +#undef PRINT_FLAG } static void sseu_dump(const struct sseu_dev_info *sseu, struct drm_printer *p) diff --git a/drivers/gpu/drm/i915/intel_device_info.h b/drivers/gpu/drm/i915/intel_device_info.h index 88f97210dc49..1caf24e2cf0b 100644 --- a/drivers/gpu/drm/i915/intel_device_info.h +++ b/drivers/gpu/drm/i915/intel_device_info.h @@ -89,35 +89,38 @@ enum intel_ppgtt { func(is_alpha_support); \ /* Keep has_* in alphabetical order */ \ func(has_64bit_reloc); \ - func(has_csr); \ - func(has_ddi); \ - func(has_dp_mst); \ func(has_reset_engine); \ - func(has_fbc); \ func(has_fpga_dbg); \ - func(has_gmch_display); \ func(has_guc); \ func(has_guc_ct); \ - func(has_hotplug); \ func(has_l3_dpf); \ func(has_llc); \ func(has_logical_ring_contexts); \ func(has_logical_ring_elsq); \ func(has_logical_ring_preemption); \ - func(has_overlay); \ func(has_pooled_eu); \ - func(has_psr); \ func(has_rc6); \ func(has_rc6p); \ func(has_runtime_pm); \ func(has_snoop); \ func(has_coherent_ggtt); \ func(unfenced_needs_alignment); \ + func(hws_needs_physical); + +#define DEV_INFO_DISPLAY_FOR_EACH_FLAG(func) \ + /* Keep in alphabetical order */ \ func(cursor_needs_physical); \ - func(hws_needs_physical); \ + func(has_csr); \ + func(has_ddi); \ + func(has_dp_mst); \ + func(has_fbc); \ + func(has_gmch_display); \ + func(has_hotplug); \ + func(has_ipc); \ + func(has_overlay); \ + func(has_psr); \ func(overlay_needs_physical); \ - func(supports_tv); \ - func(has_ipc); + func(supports_tv); #define GEN_MAX_SLICES (6) /* CNL upper bound */ #define GEN_MAX_SUBSLICES (8) /* ICL upper bound */ @@ -172,6 +175,13 @@ struct intel_device_info { #define DEFINE_FLAG(name) u8 name:1 DEV_INFO_FOR_EACH_FLAG(DEFINE_FLAG); #undef DEFINE_FLAG + + struct { +#define DEFINE_FLAG(name) u8 name:1 + DEV_INFO_DISPLAY_FOR_EACH_FLAG(DEFINE_FLAG); +#undef DEFINE_FLAG + } display; + u16 ddb_size; /* in blocks */ /* Register offsets for the various display pipes and transcoders */ diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index aa5f89f26cf5..a2584f977ab1 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -9717,7 +9717,7 @@ static u32 intel_cursor_base(const struct intel_plane_state *plane_state) const struct drm_i915_gem_object *obj = intel_fb_obj(fb); u32 base; - if (INTEL_INFO(dev_priv)->cursor_needs_physical) + if (INTEL_INFO(dev_priv)->display.cursor_needs_physical) base = obj->phys_handle->busaddr; else base = intel_plane_ggtt_offset(plane_state); @@ -13303,7 +13303,7 @@ static int intel_plane_pin_fb(struct intel_plane_state *plane_state) struct i915_vma *vma; if (plane->id == PLANE_CURSOR && - INTEL_INFO(dev_priv)->cursor_needs_physical) { + INTEL_INFO(dev_priv)->display.cursor_needs_physical) { struct drm_i915_gem_object *obj = intel_fb_obj(fb); const int align = intel_cursor_alignment(dev_priv); int err; diff --git a/drivers/gpu/drm/i915/intel_fbc.c b/drivers/gpu/drm/i915/intel_fbc.c index 14cbaf4a0e93..f23570c44323 100644 --- a/drivers/gpu/drm/i915/intel_fbc.c +++ b/drivers/gpu/drm/i915/intel_fbc.c @@ -1309,7 +1309,7 @@ void intel_fbc_init(struct drm_i915_private *dev_priv) fbc->active = false; if (need_fbc_vtd_wa(dev_priv)) - mkwrite_device_info(dev_priv)->has_fbc = false; + mkwrite_device_info(dev_priv)->display.has_fbc = false; i915_modparams.enable_fbc = intel_sanitize_fbc_option(dev_priv); DRM_DEBUG_KMS("Sanitized enable_fbc value: %d\n", From 0ce611c906bff3d6c3d72f3d5d8de79ea0490fa0 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Tue, 20 Nov 2018 20:24:39 +0000 Subject: [PATCH 86/98] drm/i915/dp: Fix inconsistent indenting Always show the FEC capability as it is initialised to 0 before error. Fixing, drivers/gpu/drm/i915/intel_dp.c:3846 intel_dp_get_dsc_sink_cap() warn: inconsistent indenting Fixes: 08cadae8e157 ("i915/dp/fec: Cache the FEC_CAPABLE DPCD register") Signed-off-by: Chris Wilson Cc: Jani Nikula Cc: Ville Syrjala Cc: Manasi Navare Cc: Anusha Srivatsa Reviewed-by: Anusha Srivatsa Link: https://patchwork.freedesktop.org/patch/msgid/20181120202439.13017-2-chris@chris-wilson.co.uk --- drivers/gpu/drm/i915/intel_dp.c | 15 +++++++-------- 1 file changed, 7 insertions(+), 8 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index a6907a1761ab..fdd2cbc56fa3 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -4026,15 +4026,14 @@ static void intel_dp_get_dsc_sink_cap(struct intel_dp *intel_dp) DRM_DEBUG_KMS("DSC DPCD: %*ph\n", (int)sizeof(intel_dp->dsc_dpcd), intel_dp->dsc_dpcd); - /* FEC is supported only on DP 1.4 */ - if (!intel_dp_is_edp(intel_dp)) { - if (drm_dp_dpcd_readb(&intel_dp->aux, DP_FEC_CAPABILITY, - &intel_dp->fec_capable) < 0) - DRM_ERROR("Failed to read FEC DPCD register\n"); - DRM_DEBUG_KMS("FEC CAPABILITY: %x\n", - intel_dp->fec_capable); - } + /* FEC is supported only on DP 1.4 */ + if (!intel_dp_is_edp(intel_dp) && + drm_dp_dpcd_readb(&intel_dp->aux, DP_FEC_CAPABILITY, + &intel_dp->fec_capable) < 0) + DRM_ERROR("Failed to read FEC DPCD register\n"); + + DRM_DEBUG_KMS("FEC CAPABILITY: %x\n", intel_dp->fec_capable); } } From 3800960afe158fa3d4c622774eaf59a9b3960a82 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Mon, 3 Dec 2018 11:36:55 +0000 Subject: [PATCH 87/98] drm/i915: Complete the fences as they are cancelled due to wedging We inspect the requests under the assumption that they will be marked as completed when they are removed from the queue. Currently however, in the process of wedging the requests will be removed from the queue before they are completed, so rearrange the code to complete the fences before the locks are dropped. <1>[ 354.473346] BUG: unable to handle kernel NULL pointer dereference at 0000000000000250 <6>[ 354.473363] PGD 0 P4D 0 <4>[ 354.473370] Oops: 0000 [#1] PREEMPT SMP PTI <4>[ 354.473380] CPU: 0 PID: 4470 Comm: gem_eio Tainted: G U 4.20.0-rc4-CI-CI_DRM_5216+ #1 <4>[ 354.473393] Hardware name: Intel Corporation NUC7CJYH/NUC7JYB, BIOS JYGLKCPX.86A.0027.2018.0125.1347 01/25/2018 <4>[ 354.473480] RIP: 0010:__i915_schedule+0x311/0x5e0 [i915] <4>[ 354.473490] Code: 49 89 44 24 20 4d 89 4c 24 28 4d 89 29 44 39 b3 a0 04 00 00 7d 3a 41 8b 44 24 78 85 c0 74 13 48 8b 93 78 04 00 00 48 83 e2 fc <39> 82 50 02 00 00 79 1e 44 89 b3 a0 04 00 00 48 8d bb d0 03 00 00 <4>[ 354.473515] RSP: 0018:ffffc900001bba90 EFLAGS: 00010046 <4>[ 354.473524] RAX: 0000000000000003 RBX: ffff8882624c8008 RCX: f34a737800000000 <4>[ 354.473535] RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff8882624c8048 <4>[ 354.473545] RBP: ffffc900001bbab0 R08: 000000005963f1f1 R09: 0000000000000000 <4>[ 354.473556] R10: ffffc900001bba10 R11: ffff8882624c8060 R12: ffff88824fdd7b98 <4>[ 354.473567] R13: ffff88824fdd7bb8 R14: 0000000000000001 R15: ffff88824fdd7750 <4>[ 354.473578] FS: 00007f44b4b5b980(0000) GS:ffff888277e00000(0000) knlGS:0000000000000000 <4>[ 354.473590] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 <4>[ 354.473599] CR2: 0000000000000250 CR3: 000000026976e000 CR4: 0000000000340ef0 <4>[ 354.473611] Call Trace: <4>[ 354.473622] ? lock_acquire+0xa6/0x1c0 <4>[ 354.473677] ? i915_schedule_bump_priority+0x57/0xd0 [i915] <4>[ 354.473736] i915_schedule_bump_priority+0x72/0xd0 [i915] <4>[ 354.473792] i915_request_wait+0x4db/0x840 [i915] <4>[ 354.473804] ? get_pwq.isra.4+0x2c/0x50 <4>[ 354.473813] ? ___preempt_schedule+0x16/0x18 <4>[ 354.473824] ? wake_up_q+0x70/0x70 <4>[ 354.473831] ? wake_up_q+0x70/0x70 <4>[ 354.473882] ? gen6_rps_boost+0x118/0x120 [i915] <4>[ 354.473936] i915_gem_object_wait_fence+0x8a/0x110 [i915] <4>[ 354.473991] i915_gem_object_wait+0x113/0x500 [i915] <4>[ 354.474047] i915_gem_wait_ioctl+0x11c/0x2f0 [i915] <4>[ 354.474101] ? i915_gem_unset_wedged+0x210/0x210 [i915] <4>[ 354.474113] drm_ioctl_kernel+0x81/0xf0 <4>[ 354.474123] drm_ioctl+0x2de/0x390 <4>[ 354.474175] ? i915_gem_unset_wedged+0x210/0x210 [i915] <4>[ 354.474187] ? finish_task_switch+0x95/0x260 <4>[ 354.474197] ? lock_acquire+0xa6/0x1c0 <4>[ 354.474207] do_vfs_ioctl+0xa0/0x6e0 <4>[ 354.474217] ? __fget+0xfc/0x1e0 <4>[ 354.474225] ksys_ioctl+0x35/0x60 <4>[ 354.474233] __x64_sys_ioctl+0x11/0x20 <4>[ 354.474241] do_syscall_64+0x55/0x190 <4>[ 354.474251] entry_SYSCALL_64_after_hwframe+0x49/0xbe <4>[ 354.474260] RIP: 0033:0x7f44b3de65d7 <4>[ 354.474267] Code: b3 66 90 48 8b 05 b1 48 2d 00 64 c7 00 26 00 00 00 48 c7 c0 ff ff ff ff c3 66 2e 0f 1f 84 00 00 00 00 00 b8 10 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 81 48 2d 00 f7 d8 64 89 01 48 <4>[ 354.474293] RSP: 002b:00007fff974948e8 EFLAGS: 00000246 ORIG_RAX: 0000000000000010 <4>[ 354.474305] RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f44b3de65d7 <4>[ 354.474316] RDX: 00007fff97494940 RSI: 00000000c010646c RDI: 0000000000000007 <4>[ 354.474327] RBP: 00007fff97494940 R08: 0000000000000000 R09: 00007f44b40bbc40 <4>[ 354.474337] R10: 0000000000000000 R11: 0000000000000246 R12: 00000000c010646c <4>[ 354.474348] R13: 0000000000000007 R14: 0000000000000000 R15: 0000000000000000 v2: Avoid floating requests. v3: Can't call dma_fence_signal() under the timeline lock! v4: Can't call dma_fence_signal() from inside another fence either. Signed-off-by: Chris Wilson Reviewed-by: Tvrtko Ursulin Link: https://patchwork.freedesktop.org/patch/msgid/20181203113701.12106-2-chris@chris-wilson.co.uk --- drivers/gpu/drm/i915/i915_gem.c | 54 +++++-------------------- drivers/gpu/drm/i915/intel_lrc.c | 11 ++++- drivers/gpu/drm/i915/intel_ringbuffer.c | 13 +++++- 3 files changed, 30 insertions(+), 48 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index c55b1f75c980..834240a9b262 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -3308,16 +3308,6 @@ void i915_gem_reset_finish(struct drm_i915_private *dev_priv) } static void nop_submit_request(struct i915_request *request) -{ - GEM_TRACE("%s fence %llx:%d -> -EIO\n", - request->engine->name, - request->fence.context, request->fence.seqno); - dma_fence_set_error(&request->fence, -EIO); - - i915_request_submit(request); -} - -static void nop_complete_submit_request(struct i915_request *request) { unsigned long flags; @@ -3354,57 +3344,33 @@ void i915_gem_set_wedged(struct drm_i915_private *i915) * rolling the global seqno forward (since this would complete requests * for which we haven't set the fence error to EIO yet). */ - for_each_engine(engine, i915, id) { + for_each_engine(engine, i915, id) i915_gem_reset_prepare_engine(engine); - engine->submit_request = nop_submit_request; - engine->schedule = NULL; - } - i915->caps.scheduler = 0; - /* Even if the GPU reset fails, it should still stop the engines */ if (INTEL_GEN(i915) >= 5) intel_gpu_reset(i915, ALL_ENGINES); - /* - * Make sure no one is running the old callback before we proceed with - * cancelling requests and resetting the completion tracking. Otherwise - * we might submit a request to the hardware which never completes. - */ - synchronize_rcu(); - for_each_engine(engine, i915, id) { - /* Mark all executing requests as skipped */ - engine->cancel_requests(engine); - - /* - * Only once we've force-cancelled all in-flight requests can we - * start to complete all requests. - */ - engine->submit_request = nop_complete_submit_request; + engine->submit_request = nop_submit_request; + engine->schedule = NULL; } + i915->caps.scheduler = 0; /* * Make sure no request can slip through without getting completed by * either this call here to intel_engine_init_global_seqno, or the one - * in nop_complete_submit_request. + * in nop_submit_request. */ synchronize_rcu(); + /* Mark all executing requests as skipped */ + for_each_engine(engine, i915, id) + engine->cancel_requests(engine); + for_each_engine(engine, i915, id) { - unsigned long flags; - - /* - * Mark all pending requests as complete so that any concurrent - * (lockless) lookup doesn't try and wait upon the request as we - * reset it. - */ - spin_lock_irqsave(&engine->timeline.lock, flags); - intel_engine_init_global_seqno(engine, - intel_engine_last_submit(engine)); - spin_unlock_irqrestore(&engine->timeline.lock, flags); - i915_gem_reset_finish_engine(engine); + intel_engine_wakeup(engine); } out: diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c index 1f004683b777..87d42a2b9400 100644 --- a/drivers/gpu/drm/i915/intel_lrc.c +++ b/drivers/gpu/drm/i915/intel_lrc.c @@ -820,8 +820,11 @@ static void execlists_cancel_requests(struct intel_engine_cs *engine) /* Mark all executing requests as skipped. */ list_for_each_entry(rq, &engine->timeline.requests, link) { GEM_BUG_ON(!rq->global_seqno); - if (!i915_request_completed(rq)) - dma_fence_set_error(&rq->fence, -EIO); + + if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &rq->fence.flags)) + continue; + + dma_fence_set_error(&rq->fence, -EIO); } /* Flush the queued requests to the timeline list (for retiring). */ @@ -841,6 +844,10 @@ static void execlists_cancel_requests(struct intel_engine_cs *engine) kmem_cache_free(engine->i915->priorities, p); } + intel_write_status_page(engine, + I915_GEM_HWS_INDEX, + intel_engine_last_submit(engine)); + /* Remaining _unready_ requests will be nop'ed when submitted */ execlists->queue_priority = INT_MIN; diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index d81eaf5f6b3e..81b10d85b738 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -755,9 +755,18 @@ static void cancel_requests(struct intel_engine_cs *engine) /* Mark all submitted requests as skipped. */ list_for_each_entry(request, &engine->timeline.requests, link) { GEM_BUG_ON(!request->global_seqno); - if (!i915_request_completed(request)) - dma_fence_set_error(&request->fence, -EIO); + + if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, + &request->fence.flags)) + continue; + + dma_fence_set_error(&request->fence, -EIO); } + + intel_write_status_page(engine, + I915_GEM_HWS_INDEX, + intel_engine_last_submit(engine)); + /* Remaining _unready_ requests will be nop'ed when submitted */ spin_unlock_irqrestore(&engine->timeline.lock, flags); From 635b3bc6f514f7fd927309a51ecf12b6f9afcac1 Mon Sep 17 00:00:00 2001 From: Jonathan Gray Date: Thu, 29 Nov 2018 12:30:51 +1100 Subject: [PATCH 88/98] drm/i915: change i915_sw_fence license to MIT Change the license of the i915_sw_fence files to MIT matching most of the other i915 files. This makes it possible to use them in a new port of i915 to OpenBSD. Besides some mechanical tree wide changes Chris Wilson is the sole author of these files with Intel holding the copyright. Intel's legal team have given permission to change the license according to Joonas Lahtinen. v2: expand commit message and note permission from Intel legal Signed-off-by: Jonathan Gray Cc: Jani Nikula Cc: Joonas Lahtinen Cc: Rodrigo Vivi Reviewed-by: Chris Wilson Reviewed-by: Joonas Lahtinen Signed-off-by: Chris Wilson Link: https://patchwork.freedesktop.org/patch/msgid/20181129013051.17525-1-jsg@jsg.id.au --- drivers/gpu/drm/i915/i915_sw_fence.c | 7 ++----- drivers/gpu/drm/i915/i915_sw_fence.h | 5 ++--- 2 files changed, 4 insertions(+), 8 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_sw_fence.c b/drivers/gpu/drm/i915/i915_sw_fence.c index 6dbeed079ae5..fc2eeab823b7 100644 --- a/drivers/gpu/drm/i915/i915_sw_fence.c +++ b/drivers/gpu/drm/i915/i915_sw_fence.c @@ -1,10 +1,7 @@ /* - * (C) Copyright 2016 Intel Corporation + * SPDX-License-Identifier: MIT * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License - * as published by the Free Software Foundation; version 2 - * of the License. + * (C) Copyright 2016 Intel Corporation */ #include diff --git a/drivers/gpu/drm/i915/i915_sw_fence.h b/drivers/gpu/drm/i915/i915_sw_fence.h index fe2ef4dadfc6..0e055ea0179f 100644 --- a/drivers/gpu/drm/i915/i915_sw_fence.h +++ b/drivers/gpu/drm/i915/i915_sw_fence.h @@ -1,10 +1,9 @@ /* + * SPDX-License-Identifier: MIT + * * i915_sw_fence.h - library routines for N:M synchronisation points * * Copyright (C) 2016 Intel Corporation - * - * This file is released under the GPLv2. - * */ #ifndef _I915_SW_FENCE_H_ From 25d140faaa25f728159eb8c304eae53d88a7f14e Mon Sep 17 00:00:00 2001 From: Tvrtko Ursulin Date: Mon, 3 Dec 2018 13:33:19 +0000 Subject: [PATCH 89/98] drm/i915: Record GT workarounds in a list To enable later verification of GT workaround state at various stages of driver lifetime, we record the list of applicable ones per platforms to a list, from which they are also applied. The added data structure is a simple array of register, mask and value items, which is allocated on demand as workarounds are added to the list. This is a temporary implementation which later in the series gets fused with the existing per context workaround list handling. It is separated at this stage since the following patch fixes a bug which needs to be as easy to backport as possible. Also, since in the following patch we will be adding a new class of workarounds (per engine) which can be applied from interrupt context, we straight away make the provision for safe read-modify-write cycle. v2: * Change dev_priv to i915 along the init path. (Chris Wilson) * API rename. (Chris Wilson) v3: * Remove explicit list size tracking in favour of growing the allocation in power of two chunks. (Chris Wilson) v4: Chris Wilson: * Change wa_list_finish to early return. * Copy workarounds using the compiler for static checking. * Do not bother zeroing unused entries. * Re-order struct i915_wa_list. v5: * kmalloc_array. * Whitespace cleanup. Signed-off-by: Tvrtko Ursulin Reviewed-by: Chris Wilson Link: https://patchwork.freedesktop.org/patch/msgid/20181203133319.10174-1-tvrtko.ursulin@linux.intel.com --- drivers/gpu/drm/i915/i915_drv.c | 1 + drivers/gpu/drm/i915/i915_drv.h | 2 + drivers/gpu/drm/i915/i915_gem.c | 4 +- drivers/gpu/drm/i915/intel_workarounds.c | 492 +++++++++++++++-------- drivers/gpu/drm/i915/intel_workarounds.h | 23 +- 5 files changed, 354 insertions(+), 168 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c index c2a3ba7e09e0..6f497cf31ffc 100644 --- a/drivers/gpu/drm/i915/i915_drv.c +++ b/drivers/gpu/drm/i915/i915_drv.c @@ -1467,6 +1467,7 @@ static int i915_driver_init_hw(struct drm_i915_private *dev_priv) intel_uncore_sanitize(dev_priv); + intel_gt_init_workarounds(dev_priv); i915_gem_load_init_fences(dev_priv); /* On the 945G/GM, the chipset reports the MSI capability on the diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 12b8476b09a2..d725390d5a48 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -69,6 +69,7 @@ #include "intel_ringbuffer.h" #include "intel_uncore.h" #include "intel_wopcm.h" +#include "intel_workarounds.h" #include "intel_uc.h" #include "i915_gem.h" @@ -1653,6 +1654,7 @@ struct drm_i915_private { int dpio_phy_iosf_port[I915_NUM_PHYS_VLV]; struct i915_workarounds workarounds; + struct i915_wa_list gt_wa_list; struct i915_frontbuffer_tracking fb_tracking; diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 834240a9b262..ad5a115e47b6 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -5300,7 +5300,7 @@ int i915_gem_init_hw(struct drm_i915_private *dev_priv) I915_WRITE(MI_PREDICATE_RESULT_2, IS_HSW_GT3(dev_priv) ? LOWER_SLICE_ENABLED : LOWER_SLICE_DISABLED); - intel_gt_workarounds_apply(dev_priv); + intel_gt_apply_workarounds(dev_priv); i915_gem_init_swizzling(dev_priv); @@ -5672,6 +5672,8 @@ void i915_gem_fini(struct drm_i915_private *dev_priv) i915_gem_contexts_fini(dev_priv); mutex_unlock(&dev_priv->drm.struct_mutex); + intel_wa_list_free(&dev_priv->gt_wa_list); + intel_cleanup_gt_powersave(dev_priv); intel_uc_fini_misc(dev_priv); diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c index e5cd6c6c66c3..c3256ab2c411 100644 --- a/drivers/gpu/drm/i915/intel_workarounds.c +++ b/drivers/gpu/drm/i915/intel_workarounds.c @@ -48,6 +48,20 @@ * - Public functions to init or apply the given workaround type. */ +static void wa_init_start(struct i915_wa_list *wal, const char *name) +{ + wal->name = name; +} + +static void wa_init_finish(struct i915_wa_list *wal) +{ + if (!wal->count) + return; + + DRM_DEBUG_DRIVER("Initialized %u %s workarounds\n", + wal->count, wal->name); +} + static void wa_add(struct drm_i915_private *i915, i915_reg_t reg, const u32 mask, const u32 val) { @@ -575,160 +589,239 @@ int intel_ctx_workarounds_emit(struct i915_request *rq) return 0; } -static void bdw_gt_workarounds_apply(struct drm_i915_private *dev_priv) +static void +wal_add(struct i915_wa_list *wal, const struct i915_wa *wa) { + const unsigned int grow = 1 << 4; + + GEM_BUG_ON(!is_power_of_2(grow)); + + if (IS_ALIGNED(wal->count, grow)) { /* Either uninitialized or full. */ + struct i915_wa *list; + + list = kmalloc_array(ALIGN(wal->count + 1, grow), sizeof(*wa), + GFP_KERNEL); + if (!list) { + DRM_ERROR("No space for workaround init!\n"); + return; + } + + if (wal->list) + memcpy(list, wal->list, sizeof(*wa) * wal->count); + + wal->list = list; + } + + wal->list[wal->count++] = *wa; } -static void chv_gt_workarounds_apply(struct drm_i915_private *dev_priv) +static void +wa_masked_en(struct i915_wa_list *wal, i915_reg_t reg, u32 val) { + struct i915_wa wa = { + .reg = reg, + .mask = val, + .val = _MASKED_BIT_ENABLE(val) + }; + + wal_add(wal, &wa); } -static void gen9_gt_workarounds_apply(struct drm_i915_private *dev_priv) +static void +wa_write_masked_or(struct i915_wa_list *wal, i915_reg_t reg, u32 mask, + u32 val) { + struct i915_wa wa = { + .reg = reg, + .mask = mask, + .val = val + }; + + wal_add(wal, &wa); +} + +static void +wa_write(struct i915_wa_list *wal, i915_reg_t reg, u32 val) +{ + wa_write_masked_or(wal, reg, ~0, val); +} + +static void +wa_write_or(struct i915_wa_list *wal, i915_reg_t reg, u32 val) +{ + wa_write_masked_or(wal, reg, val, val); +} + +static void gen9_gt_workarounds_init(struct drm_i915_private *i915) +{ + struct i915_wa_list *wal = &i915->gt_wa_list; + /* WaContextSwitchWithConcurrentTLBInvalidate:skl,bxt,kbl,glk,cfl */ - I915_WRITE(GEN9_CSFE_CHICKEN1_RCS, - _MASKED_BIT_ENABLE(GEN9_PREEMPT_GPGPU_SYNC_SWITCH_DISABLE)); + wa_masked_en(wal, + GEN9_CSFE_CHICKEN1_RCS, + GEN9_PREEMPT_GPGPU_SYNC_SWITCH_DISABLE); + /* WaEnableLbsSlaRetryTimerDecrement:skl,bxt,kbl,glk,cfl */ - I915_WRITE(BDW_SCRATCH1, I915_READ(BDW_SCRATCH1) | - GEN9_LBS_SLA_RETRY_TIMER_DECREMENT_ENABLE); + wa_write_or(wal, + BDW_SCRATCH1, + GEN9_LBS_SLA_RETRY_TIMER_DECREMENT_ENABLE); /* WaDisableKillLogic:bxt,skl,kbl */ - if (!IS_COFFEELAKE(dev_priv)) - I915_WRITE(GAM_ECOCHK, I915_READ(GAM_ECOCHK) | - ECOCHK_DIS_TLB); + if (!IS_COFFEELAKE(i915)) + wa_write_or(wal, + GAM_ECOCHK, + ECOCHK_DIS_TLB); - if (HAS_LLC(dev_priv)) { + if (HAS_LLC(i915)) { /* WaCompressedResourceSamplerPbeMediaNewHashMode:skl,kbl * * Must match Display Engine. See * WaCompressedResourceDisplayNewHashMode. */ - I915_WRITE(MMCD_MISC_CTRL, - I915_READ(MMCD_MISC_CTRL) | - MMCD_PCLA | - MMCD_HOTSPOT_EN); + wa_write_or(wal, + MMCD_MISC_CTRL, + MMCD_PCLA | MMCD_HOTSPOT_EN); } /* WaDisableHDCInvalidation:skl,bxt,kbl,cfl */ - I915_WRITE(GAM_ECOCHK, I915_READ(GAM_ECOCHK) | - BDW_DISABLE_HDC_INVALIDATION); + wa_write_or(wal, + GAM_ECOCHK, + BDW_DISABLE_HDC_INVALIDATION); /* WaProgramL3SqcReg1DefaultForPerf:bxt,glk */ - if (IS_GEN9_LP(dev_priv)) { - u32 val = I915_READ(GEN8_L3SQCREG1); - - val &= ~L3_PRIO_CREDITS_MASK; - val |= L3_GENERAL_PRIO_CREDITS(62) | L3_HIGH_PRIO_CREDITS(2); - I915_WRITE(GEN8_L3SQCREG1, val); - } + if (IS_GEN9_LP(i915)) + wa_write_masked_or(wal, + GEN8_L3SQCREG1, + L3_PRIO_CREDITS_MASK, + L3_GENERAL_PRIO_CREDITS(62) | + L3_HIGH_PRIO_CREDITS(2)); /* WaOCLCoherentLineFlush:skl,bxt,kbl,cfl */ - I915_WRITE(GEN8_L3SQCREG4, - I915_READ(GEN8_L3SQCREG4) | GEN8_LQSC_FLUSH_COHERENT_LINES); + wa_write_or(wal, + GEN8_L3SQCREG4, + GEN8_LQSC_FLUSH_COHERENT_LINES); /* WaEnablePreemptionGranularityControlByUMD:skl,bxt,kbl,cfl,[cnl] */ - I915_WRITE(GEN7_FF_SLICE_CS_CHICKEN1, - _MASKED_BIT_ENABLE(GEN9_FFSC_PERCTX_PREEMPT_CTRL)); + wa_masked_en(wal, + GEN7_FF_SLICE_CS_CHICKEN1, + GEN9_FFSC_PERCTX_PREEMPT_CTRL); } -static void skl_gt_workarounds_apply(struct drm_i915_private *dev_priv) +static void skl_gt_workarounds_init(struct drm_i915_private *i915) { - gen9_gt_workarounds_apply(dev_priv); + struct i915_wa_list *wal = &i915->gt_wa_list; + + gen9_gt_workarounds_init(i915); /* WaEnableGapsTsvCreditFix:skl */ - I915_WRITE(GEN8_GARBCNTL, - I915_READ(GEN8_GARBCNTL) | GEN9_GAPS_TSV_CREDIT_DISABLE); + wa_write_or(wal, + GEN8_GARBCNTL, + GEN9_GAPS_TSV_CREDIT_DISABLE); /* WaDisableGafsUnitClkGating:skl */ - I915_WRITE(GEN7_UCGCTL4, - I915_READ(GEN7_UCGCTL4) | GEN8_EU_GAUNIT_CLOCK_GATE_DISABLE); + wa_write_or(wal, + GEN7_UCGCTL4, + GEN8_EU_GAUNIT_CLOCK_GATE_DISABLE); /* WaInPlaceDecompressionHang:skl */ - if (IS_SKL_REVID(dev_priv, SKL_REVID_H0, REVID_FOREVER)) - I915_WRITE(GEN9_GAMT_ECO_REG_RW_IA, - I915_READ(GEN9_GAMT_ECO_REG_RW_IA) | - GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); + if (IS_SKL_REVID(i915, SKL_REVID_H0, REVID_FOREVER)) + wa_write_or(wal, + GEN9_GAMT_ECO_REG_RW_IA, + GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); } -static void bxt_gt_workarounds_apply(struct drm_i915_private *dev_priv) +static void bxt_gt_workarounds_init(struct drm_i915_private *i915) { - gen9_gt_workarounds_apply(dev_priv); + struct i915_wa_list *wal = &i915->gt_wa_list; + + gen9_gt_workarounds_init(i915); /* WaDisablePooledEuLoadBalancingFix:bxt */ - I915_WRITE(FF_SLICE_CS_CHICKEN2, - _MASKED_BIT_ENABLE(GEN9_POOLED_EU_LOAD_BALANCING_FIX_DISABLE)); + wa_masked_en(wal, + FF_SLICE_CS_CHICKEN2, + GEN9_POOLED_EU_LOAD_BALANCING_FIX_DISABLE); /* WaInPlaceDecompressionHang:bxt */ - I915_WRITE(GEN9_GAMT_ECO_REG_RW_IA, - I915_READ(GEN9_GAMT_ECO_REG_RW_IA) | - GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); + wa_write_or(wal, + GEN9_GAMT_ECO_REG_RW_IA, + GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); } -static void kbl_gt_workarounds_apply(struct drm_i915_private *dev_priv) +static void kbl_gt_workarounds_init(struct drm_i915_private *i915) { - gen9_gt_workarounds_apply(dev_priv); + struct i915_wa_list *wal = &i915->gt_wa_list; + + gen9_gt_workarounds_init(i915); /* WaEnableGapsTsvCreditFix:kbl */ - I915_WRITE(GEN8_GARBCNTL, - I915_READ(GEN8_GARBCNTL) | GEN9_GAPS_TSV_CREDIT_DISABLE); + wa_write_or(wal, + GEN8_GARBCNTL, + GEN9_GAPS_TSV_CREDIT_DISABLE); /* WaDisableDynamicCreditSharing:kbl */ - if (IS_KBL_REVID(dev_priv, 0, KBL_REVID_B0)) - I915_WRITE(GAMT_CHKN_BIT_REG, - I915_READ(GAMT_CHKN_BIT_REG) | - GAMT_CHKN_DISABLE_DYNAMIC_CREDIT_SHARING); + if (IS_KBL_REVID(i915, 0, KBL_REVID_B0)) + wa_write_or(wal, + GAMT_CHKN_BIT_REG, + GAMT_CHKN_DISABLE_DYNAMIC_CREDIT_SHARING); /* WaDisableGafsUnitClkGating:kbl */ - I915_WRITE(GEN7_UCGCTL4, - I915_READ(GEN7_UCGCTL4) | GEN8_EU_GAUNIT_CLOCK_GATE_DISABLE); + wa_write_or(wal, + GEN7_UCGCTL4, + GEN8_EU_GAUNIT_CLOCK_GATE_DISABLE); /* WaInPlaceDecompressionHang:kbl */ - I915_WRITE(GEN9_GAMT_ECO_REG_RW_IA, - I915_READ(GEN9_GAMT_ECO_REG_RW_IA) | - GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); + wa_write_or(wal, + GEN9_GAMT_ECO_REG_RW_IA, + GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); /* WaKBLVECSSemaphoreWaitPoll:kbl */ - if (IS_KBL_REVID(dev_priv, KBL_REVID_A0, KBL_REVID_E0)) { + if (IS_KBL_REVID(i915, KBL_REVID_A0, KBL_REVID_E0)) { struct intel_engine_cs *engine; unsigned int tmp; - for_each_engine(engine, dev_priv, tmp) { + for_each_engine(engine, i915, tmp) { if (engine->id == RCS) continue; - I915_WRITE(RING_SEMA_WAIT_POLL(engine->mmio_base), 1); + wa_write(wal, + RING_SEMA_WAIT_POLL(engine->mmio_base), + 1); } } } -static void glk_gt_workarounds_apply(struct drm_i915_private *dev_priv) +static void glk_gt_workarounds_init(struct drm_i915_private *i915) { - gen9_gt_workarounds_apply(dev_priv); + gen9_gt_workarounds_init(i915); } -static void cfl_gt_workarounds_apply(struct drm_i915_private *dev_priv) +static void cfl_gt_workarounds_init(struct drm_i915_private *i915) { - gen9_gt_workarounds_apply(dev_priv); + struct i915_wa_list *wal = &i915->gt_wa_list; + + gen9_gt_workarounds_init(i915); /* WaEnableGapsTsvCreditFix:cfl */ - I915_WRITE(GEN8_GARBCNTL, - I915_READ(GEN8_GARBCNTL) | GEN9_GAPS_TSV_CREDIT_DISABLE); + wa_write_or(wal, + GEN8_GARBCNTL, + GEN9_GAPS_TSV_CREDIT_DISABLE); /* WaDisableGafsUnitClkGating:cfl */ - I915_WRITE(GEN7_UCGCTL4, - I915_READ(GEN7_UCGCTL4) | GEN8_EU_GAUNIT_CLOCK_GATE_DISABLE); + wa_write_or(wal, + GEN7_UCGCTL4, + GEN8_EU_GAUNIT_CLOCK_GATE_DISABLE); /* WaInPlaceDecompressionHang:cfl */ - I915_WRITE(GEN9_GAMT_ECO_REG_RW_IA, - I915_READ(GEN9_GAMT_ECO_REG_RW_IA) | - GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); + wa_write_or(wal, + GEN9_GAMT_ECO_REG_RW_IA, + GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); } static void wa_init_mcr(struct drm_i915_private *dev_priv) { const struct sseu_dev_info *sseu = &(INTEL_INFO(dev_priv)->sseu); - u32 mcr; + struct i915_wa_list *wal = &dev_priv->gt_wa_list; u32 mcr_slice_subslice_mask; /* @@ -765,8 +858,6 @@ static void wa_init_mcr(struct drm_i915_private *dev_priv) WARN_ON((enabled_mask & disabled_mask) != enabled_mask); } - mcr = I915_READ(GEN8_MCR_SELECTOR); - if (INTEL_GEN(dev_priv) >= 11) mcr_slice_subslice_mask = GEN11_MCR_SLICE_MASK | GEN11_MCR_SUBSLICE_MASK; @@ -784,156 +875,225 @@ static void wa_init_mcr(struct drm_i915_private *dev_priv) * occasions, such as INSTDONE, where this value is dependent * on s/ss combo, the read should be done with read_subslice_reg. */ - mcr &= ~mcr_slice_subslice_mask; - mcr |= intel_calculate_mcr_s_ss_select(dev_priv); - I915_WRITE(GEN8_MCR_SELECTOR, mcr); + wa_write_masked_or(wal, + GEN8_MCR_SELECTOR, + mcr_slice_subslice_mask, + intel_calculate_mcr_s_ss_select(dev_priv)); } -static void cnl_gt_workarounds_apply(struct drm_i915_private *dev_priv) +static void cnl_gt_workarounds_init(struct drm_i915_private *i915) { - wa_init_mcr(dev_priv); + struct i915_wa_list *wal = &i915->gt_wa_list; + + wa_init_mcr(i915); /* WaDisableI2mCycleOnWRPort:cnl (pre-prod) */ - if (IS_CNL_REVID(dev_priv, CNL_REVID_B0, CNL_REVID_B0)) - I915_WRITE(GAMT_CHKN_BIT_REG, - I915_READ(GAMT_CHKN_BIT_REG) | - GAMT_CHKN_DISABLE_I2M_CYCLE_ON_WR_PORT); + if (IS_CNL_REVID(i915, CNL_REVID_B0, CNL_REVID_B0)) + wa_write_or(wal, + GAMT_CHKN_BIT_REG, + GAMT_CHKN_DISABLE_I2M_CYCLE_ON_WR_PORT); /* WaInPlaceDecompressionHang:cnl */ - I915_WRITE(GEN9_GAMT_ECO_REG_RW_IA, - I915_READ(GEN9_GAMT_ECO_REG_RW_IA) | - GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); + wa_write_or(wal, + GEN9_GAMT_ECO_REG_RW_IA, + GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); /* WaEnablePreemptionGranularityControlByUMD:cnl */ - I915_WRITE(GEN7_FF_SLICE_CS_CHICKEN1, - _MASKED_BIT_ENABLE(GEN9_FFSC_PERCTX_PREEMPT_CTRL)); + wa_masked_en(wal, + GEN7_FF_SLICE_CS_CHICKEN1, + GEN9_FFSC_PERCTX_PREEMPT_CTRL); } -static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv) +static void icl_gt_workarounds_init(struct drm_i915_private *i915) { - wa_init_mcr(dev_priv); + struct i915_wa_list *wal = &i915->gt_wa_list; + + wa_init_mcr(i915); /* This is not an Wa. Enable for better image quality */ - I915_WRITE(_3D_CHICKEN3, - _MASKED_BIT_ENABLE(_3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLE)); + wa_masked_en(wal, + _3D_CHICKEN3, + _3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLE); /* WaInPlaceDecompressionHang:icl */ - I915_WRITE(GEN9_GAMT_ECO_REG_RW_IA, - I915_READ(GEN9_GAMT_ECO_REG_RW_IA) | - GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); + wa_write_or(wal, + GEN9_GAMT_ECO_REG_RW_IA, + GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); /* WaPipelineFlushCoherentLines:icl */ - I915_WRITE(GEN8_L3SQCREG4, - I915_READ(GEN8_L3SQCREG4) | - GEN8_LQSC_FLUSH_COHERENT_LINES); + wa_write_or(wal, + GEN8_L3SQCREG4, + GEN8_LQSC_FLUSH_COHERENT_LINES); /* Wa_1405543622:icl * Formerly known as WaGAPZPriorityScheme */ - I915_WRITE(GEN8_GARBCNTL, - I915_READ(GEN8_GARBCNTL) | - GEN11_ARBITRATION_PRIO_ORDER_MASK); + wa_write_or(wal, + GEN8_GARBCNTL, + GEN11_ARBITRATION_PRIO_ORDER_MASK); /* Wa_1604223664:icl * Formerly known as WaL3BankAddressHashing */ - I915_WRITE(GEN8_GARBCNTL, - (I915_READ(GEN8_GARBCNTL) & ~GEN11_HASH_CTRL_EXCL_MASK) | - GEN11_HASH_CTRL_EXCL_BIT0); - I915_WRITE(GEN11_GLBLINVL, - (I915_READ(GEN11_GLBLINVL) & ~GEN11_BANK_HASH_ADDR_EXCL_MASK) | - GEN11_BANK_HASH_ADDR_EXCL_BIT0); + wa_write_masked_or(wal, + GEN8_GARBCNTL, + GEN11_HASH_CTRL_EXCL_MASK, + GEN11_HASH_CTRL_EXCL_BIT0); + wa_write_masked_or(wal, + GEN11_GLBLINVL, + GEN11_BANK_HASH_ADDR_EXCL_MASK, + GEN11_BANK_HASH_ADDR_EXCL_BIT0); /* WaModifyGamTlbPartitioning:icl */ - I915_WRITE(GEN11_GACB_PERF_CTRL, - (I915_READ(GEN11_GACB_PERF_CTRL) & ~GEN11_HASH_CTRL_MASK) | - GEN11_HASH_CTRL_BIT0 | GEN11_HASH_CTRL_BIT4); + wa_write_masked_or(wal, + GEN11_GACB_PERF_CTRL, + GEN11_HASH_CTRL_MASK, + GEN11_HASH_CTRL_BIT0 | GEN11_HASH_CTRL_BIT4); /* Wa_1405733216:icl * Formerly known as WaDisableCleanEvicts */ - I915_WRITE(GEN8_L3SQCREG4, - I915_READ(GEN8_L3SQCREG4) | - GEN11_LQSC_CLEAN_EVICT_DISABLE); + wa_write_or(wal, + GEN8_L3SQCREG4, + GEN11_LQSC_CLEAN_EVICT_DISABLE); /* Wa_1405766107:icl * Formerly known as WaCL2SFHalfMaxAlloc */ - I915_WRITE(GEN11_LSN_UNSLCVC, - I915_READ(GEN11_LSN_UNSLCVC) | - GEN11_LSN_UNSLCVC_GAFS_HALF_SF_MAXALLOC | - GEN11_LSN_UNSLCVC_GAFS_HALF_CL2_MAXALLOC); + wa_write_or(wal, + GEN11_LSN_UNSLCVC, + GEN11_LSN_UNSLCVC_GAFS_HALF_SF_MAXALLOC | + GEN11_LSN_UNSLCVC_GAFS_HALF_CL2_MAXALLOC); /* Wa_220166154:icl * Formerly known as WaDisCtxReload */ - I915_WRITE(GEN8_GAMW_ECO_DEV_RW_IA, - I915_READ(GEN8_GAMW_ECO_DEV_RW_IA) | - GAMW_ECO_DEV_CTX_RELOAD_DISABLE); + wa_write_or(wal, + GEN8_GAMW_ECO_DEV_RW_IA, + GAMW_ECO_DEV_CTX_RELOAD_DISABLE); /* Wa_1405779004:icl (pre-prod) */ - if (IS_ICL_REVID(dev_priv, ICL_REVID_A0, ICL_REVID_A0)) - I915_WRITE(SLICE_UNIT_LEVEL_CLKGATE, - I915_READ(SLICE_UNIT_LEVEL_CLKGATE) | - MSCUNIT_CLKGATE_DIS); + if (IS_ICL_REVID(i915, ICL_REVID_A0, ICL_REVID_A0)) + wa_write_or(wal, + SLICE_UNIT_LEVEL_CLKGATE, + MSCUNIT_CLKGATE_DIS); /* Wa_1406680159:icl */ - I915_WRITE(SUBSLICE_UNIT_LEVEL_CLKGATE, - I915_READ(SUBSLICE_UNIT_LEVEL_CLKGATE) | - GWUNIT_CLKGATE_DIS); + wa_write_or(wal, + SUBSLICE_UNIT_LEVEL_CLKGATE, + GWUNIT_CLKGATE_DIS); /* Wa_1406838659:icl (pre-prod) */ - if (IS_ICL_REVID(dev_priv, ICL_REVID_A0, ICL_REVID_B0)) - I915_WRITE(INF_UNIT_LEVEL_CLKGATE, - I915_READ(INF_UNIT_LEVEL_CLKGATE) | - CGPSF_CLKGATE_DIS); + if (IS_ICL_REVID(i915, ICL_REVID_A0, ICL_REVID_B0)) + wa_write_or(wal, + INF_UNIT_LEVEL_CLKGATE, + CGPSF_CLKGATE_DIS); /* WaForwardProgressSoftReset:icl */ - I915_WRITE(GEN10_SCRATCH_LNCF2, - I915_READ(GEN10_SCRATCH_LNCF2) | - PMFLUSHDONE_LNICRSDROP | - PMFLUSH_GAPL3UNBLOCK | - PMFLUSHDONE_LNEBLK); + wa_write_or(wal, + GEN10_SCRATCH_LNCF2, + PMFLUSHDONE_LNICRSDROP | + PMFLUSH_GAPL3UNBLOCK | + PMFLUSHDONE_LNEBLK); /* Wa_1406463099:icl * Formerly known as WaGamTlbPendError */ - I915_WRITE(GAMT_CHKN_BIT_REG, - I915_READ(GAMT_CHKN_BIT_REG) | - GAMT_CHKN_DISABLE_L3_COH_PIPE); + wa_write_or(wal, + GAMT_CHKN_BIT_REG, + GAMT_CHKN_DISABLE_L3_COH_PIPE); /* Wa_1406609255:icl (pre-prod) */ - if (IS_ICL_REVID(dev_priv, ICL_REVID_A0, ICL_REVID_B0)) - I915_WRITE(GEN7_SARCHKMD, - I915_READ(GEN7_SARCHKMD) | - GEN7_DISABLE_DEMAND_PREFETCH | - GEN7_DISABLE_SAMPLER_PREFETCH); + if (IS_ICL_REVID(i915, ICL_REVID_A0, ICL_REVID_B0)) + wa_write_or(wal, + GEN7_SARCHKMD, + GEN7_DISABLE_DEMAND_PREFETCH | + GEN7_DISABLE_SAMPLER_PREFETCH); } -void intel_gt_workarounds_apply(struct drm_i915_private *dev_priv) +void intel_gt_init_workarounds(struct drm_i915_private *i915) { - if (INTEL_GEN(dev_priv) < 8) + struct i915_wa_list *wal = &i915->gt_wa_list; + + wa_init_start(wal, "GT"); + + if (INTEL_GEN(i915) < 8) return; - else if (IS_BROADWELL(dev_priv)) - bdw_gt_workarounds_apply(dev_priv); - else if (IS_CHERRYVIEW(dev_priv)) - chv_gt_workarounds_apply(dev_priv); - else if (IS_SKYLAKE(dev_priv)) - skl_gt_workarounds_apply(dev_priv); - else if (IS_BROXTON(dev_priv)) - bxt_gt_workarounds_apply(dev_priv); - else if (IS_KABYLAKE(dev_priv)) - kbl_gt_workarounds_apply(dev_priv); - else if (IS_GEMINILAKE(dev_priv)) - glk_gt_workarounds_apply(dev_priv); - else if (IS_COFFEELAKE(dev_priv)) - cfl_gt_workarounds_apply(dev_priv); - else if (IS_CANNONLAKE(dev_priv)) - cnl_gt_workarounds_apply(dev_priv); - else if (IS_ICELAKE(dev_priv)) - icl_gt_workarounds_apply(dev_priv); + else if (IS_BROADWELL(i915)) + return; + else if (IS_CHERRYVIEW(i915)) + return; + else if (IS_SKYLAKE(i915)) + skl_gt_workarounds_init(i915); + else if (IS_BROXTON(i915)) + bxt_gt_workarounds_init(i915); + else if (IS_KABYLAKE(i915)) + kbl_gt_workarounds_init(i915); + else if (IS_GEMINILAKE(i915)) + glk_gt_workarounds_init(i915); + else if (IS_COFFEELAKE(i915)) + cfl_gt_workarounds_init(i915); + else if (IS_CANNONLAKE(i915)) + cnl_gt_workarounds_init(i915); + else if (IS_ICELAKE(i915)) + icl_gt_workarounds_init(i915); else - MISSING_CASE(INTEL_GEN(dev_priv)); + MISSING_CASE(INTEL_GEN(i915)); + + wa_init_finish(wal); +} + +static enum forcewake_domains +wal_get_fw_for_rmw(struct drm_i915_private *dev_priv, + const struct i915_wa_list *wal) +{ + enum forcewake_domains fw = 0; + struct i915_wa *wa; + unsigned int i; + + for (i = 0, wa = wal->list; i < wal->count; i++, wa++) + fw |= intel_uncore_forcewake_for_reg(dev_priv, + wa->reg, + FW_REG_READ | + FW_REG_WRITE); + + return fw; +} + +static void +wa_list_apply(struct drm_i915_private *dev_priv, const struct i915_wa_list *wal) +{ + enum forcewake_domains fw; + unsigned long flags; + struct i915_wa *wa; + unsigned int i; + + if (!wal->count) + return; + + fw = wal_get_fw_for_rmw(dev_priv, wal); + + spin_lock_irqsave(&dev_priv->uncore.lock, flags); + intel_uncore_forcewake_get__locked(dev_priv, fw); + + for (i = 0, wa = wal->list; i < wal->count; i++, wa++) { + u32 val = I915_READ_FW(wa->reg); + + val &= ~wa->mask; + val |= wa->val; + + I915_WRITE_FW(wa->reg, val); + } + + intel_uncore_forcewake_put__locked(dev_priv, fw); + spin_unlock_irqrestore(&dev_priv->uncore.lock, flags); + + DRM_DEBUG_DRIVER("Applied %u %s workarounds\n", wal->count, wal->name); +} + +void intel_gt_apply_workarounds(struct drm_i915_private *dev_priv) +{ + wa_list_apply(dev_priv, &dev_priv->gt_wa_list); } struct whitelist { diff --git a/drivers/gpu/drm/i915/intel_workarounds.h b/drivers/gpu/drm/i915/intel_workarounds.h index b11d0623e626..263106600fdc 100644 --- a/drivers/gpu/drm/i915/intel_workarounds.h +++ b/drivers/gpu/drm/i915/intel_workarounds.h @@ -7,10 +7,31 @@ #ifndef _I915_WORKAROUNDS_H_ #define _I915_WORKAROUNDS_H_ +#include + +struct i915_wa { + i915_reg_t reg; + u32 mask; + u32 val; +}; + +struct i915_wa_list { + const char *name; + struct i915_wa *list; + unsigned int count; +}; + +static inline void intel_wa_list_free(struct i915_wa_list *wal) +{ + kfree(wal->list); + memset(wal, 0, sizeof(*wal)); +} + int intel_ctx_workarounds_init(struct drm_i915_private *dev_priv); int intel_ctx_workarounds_emit(struct i915_request *rq); -void intel_gt_workarounds_apply(struct drm_i915_private *dev_priv); +void intel_gt_init_workarounds(struct drm_i915_private *dev_priv); +void intel_gt_apply_workarounds(struct drm_i915_private *dev_priv); void intel_whitelist_workarounds_apply(struct intel_engine_cs *engine); From 4a15c75c42460252a63d30f03b4766a52945fb47 Mon Sep 17 00:00:00 2001 From: Tvrtko Ursulin Date: Mon, 3 Dec 2018 13:33:41 +0000 Subject: [PATCH 90/98] drm/i915: Introduce per-engine workarounds MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit We stopped re-applying the GT workarounds after engine reset since commit 59b449d5c82a ("drm/i915: Split out functions for different kinds of workarounds"). Issue with this is that some of the GT workarounds live in the MMIO space which gets lost during engine resets. So far the registers in 0x2xxx and 0xbxxx address range have been identified to be affected. This losing of applied workarounds has obvious negative effects and can even lead to hard system hangs (see the linked Bugzilla). Rather than just restoring this re-application, because we have also observed that it is not safe to just re-write all GT workarounds after engine resets (GPU might be live and weird hardware states can happen), we introduce a new class of per-engine workarounds and move only the affected GT workarounds over. Using the framework introduced in the previous patch, we therefore after engine reset, re-apply only the workarounds living in the affected MMIO address ranges. v2: * Move Wa_1406609255:icl to engine workarounds as well. * Rename API. (Chris Wilson) * Drop redundant IS_KABYLAKE. (Chris Wilson) * Re-order engine wa/ init so latest platforms are first. (Rodrigo Vivi) Signed-off-by: Tvrtko Ursulin Bugzilla: https://bugzilla.freedesktop.org/show_bug.cgi?id=107945 Fixes: 59b449d5c82a ("drm/i915: Split out functions for different kinds of workarounds") Cc: Mika Kuoppala Cc: Ville Syrjälä Cc: Chris Wilson Cc: Jani Nikula Cc: Joonas Lahtinen Cc: Rodrigo Vivi Cc: intel-gfx@lists.freedesktop.org Acked-by: Rodrigo Vivi Reviewed-by: Chris Wilson Link: https://patchwork.freedesktop.org/patch/msgid/20181203133341.10258-1-tvrtko.ursulin@linux.intel.com --- drivers/gpu/drm/i915/intel_engine_cs.c | 2 + drivers/gpu/drm/i915/intel_lrc.c | 4 + drivers/gpu/drm/i915/intel_ringbuffer.h | 2 + drivers/gpu/drm/i915/intel_workarounds.c | 261 ++++++++++++----------- drivers/gpu/drm/i915/intel_workarounds.h | 3 + 5 files changed, 153 insertions(+), 119 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c index 759c0fd58f8c..ef5d202e9d45 100644 --- a/drivers/gpu/drm/i915/intel_engine_cs.c +++ b/drivers/gpu/drm/i915/intel_engine_cs.c @@ -723,6 +723,8 @@ void intel_engine_cleanup_common(struct intel_engine_cs *engine) __intel_context_unpin(i915->kernel_context, engine); i915_timeline_fini(&engine->timeline); + + intel_wa_list_free(&engine->wa_list); } u64 intel_engine_get_active_head(const struct intel_engine_cs *engine) diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c index 87d42a2b9400..a5d6663ceafd 100644 --- a/drivers/gpu/drm/i915/intel_lrc.c +++ b/drivers/gpu/drm/i915/intel_lrc.c @@ -1626,6 +1626,8 @@ static bool unexpected_starting_state(struct intel_engine_cs *engine) static int gen8_init_common_ring(struct intel_engine_cs *engine) { + intel_engine_apply_workarounds(engine); + intel_mocs_init_engine(engine); intel_engine_reset_breadcrumbs(engine); @@ -2305,6 +2307,8 @@ int logical_render_ring_init(struct intel_engine_cs *engine) ret); } + intel_engine_init_workarounds(engine); + return 0; err_cleanup_common: diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h index f4988f7bc756..3abac391a739 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.h +++ b/drivers/gpu/drm/i915/intel_ringbuffer.h @@ -15,6 +15,7 @@ #include "i915_selftest.h" #include "i915_timeline.h" #include "intel_gpu_commands.h" +#include "intel_workarounds.h" struct drm_printer; struct i915_sched_attr; @@ -435,6 +436,7 @@ struct intel_engine_cs { struct intel_hw_status_page status_page; struct i915_ctx_workarounds wa_ctx; + struct i915_wa_list wa_list; struct i915_vma *scratch; u32 irq_keep_mask; /* always keep these interrupts */ diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c index c3256ab2c411..847988d15e8a 100644 --- a/drivers/gpu/drm/i915/intel_workarounds.c +++ b/drivers/gpu/drm/i915/intel_workarounds.c @@ -656,17 +656,6 @@ static void gen9_gt_workarounds_init(struct drm_i915_private *i915) { struct i915_wa_list *wal = &i915->gt_wa_list; - /* WaContextSwitchWithConcurrentTLBInvalidate:skl,bxt,kbl,glk,cfl */ - wa_masked_en(wal, - GEN9_CSFE_CHICKEN1_RCS, - GEN9_PREEMPT_GPGPU_SYNC_SWITCH_DISABLE); - - - /* WaEnableLbsSlaRetryTimerDecrement:skl,bxt,kbl,glk,cfl */ - wa_write_or(wal, - BDW_SCRATCH1, - GEN9_LBS_SLA_RETRY_TIMER_DECREMENT_ENABLE); - /* WaDisableKillLogic:bxt,skl,kbl */ if (!IS_COFFEELAKE(i915)) wa_write_or(wal, @@ -688,24 +677,6 @@ static void gen9_gt_workarounds_init(struct drm_i915_private *i915) wa_write_or(wal, GAM_ECOCHK, BDW_DISABLE_HDC_INVALIDATION); - - /* WaProgramL3SqcReg1DefaultForPerf:bxt,glk */ - if (IS_GEN9_LP(i915)) - wa_write_masked_or(wal, - GEN8_L3SQCREG1, - L3_PRIO_CREDITS_MASK, - L3_GENERAL_PRIO_CREDITS(62) | - L3_HIGH_PRIO_CREDITS(2)); - - /* WaOCLCoherentLineFlush:skl,bxt,kbl,cfl */ - wa_write_or(wal, - GEN8_L3SQCREG4, - GEN8_LQSC_FLUSH_COHERENT_LINES); - - /* WaEnablePreemptionGranularityControlByUMD:skl,bxt,kbl,cfl,[cnl] */ - wa_masked_en(wal, - GEN7_FF_SLICE_CS_CHICKEN1, - GEN9_FFSC_PERCTX_PREEMPT_CTRL); } static void skl_gt_workarounds_init(struct drm_i915_private *i915) @@ -714,11 +685,6 @@ static void skl_gt_workarounds_init(struct drm_i915_private *i915) gen9_gt_workarounds_init(i915); - /* WaEnableGapsTsvCreditFix:skl */ - wa_write_or(wal, - GEN8_GARBCNTL, - GEN9_GAPS_TSV_CREDIT_DISABLE); - /* WaDisableGafsUnitClkGating:skl */ wa_write_or(wal, GEN7_UCGCTL4, @@ -737,11 +703,6 @@ static void bxt_gt_workarounds_init(struct drm_i915_private *i915) gen9_gt_workarounds_init(i915); - /* WaDisablePooledEuLoadBalancingFix:bxt */ - wa_masked_en(wal, - FF_SLICE_CS_CHICKEN2, - GEN9_POOLED_EU_LOAD_BALANCING_FIX_DISABLE); - /* WaInPlaceDecompressionHang:bxt */ wa_write_or(wal, GEN9_GAMT_ECO_REG_RW_IA, @@ -754,11 +715,6 @@ static void kbl_gt_workarounds_init(struct drm_i915_private *i915) gen9_gt_workarounds_init(i915); - /* WaEnableGapsTsvCreditFix:kbl */ - wa_write_or(wal, - GEN8_GARBCNTL, - GEN9_GAPS_TSV_CREDIT_DISABLE); - /* WaDisableDynamicCreditSharing:kbl */ if (IS_KBL_REVID(i915, 0, KBL_REVID_B0)) wa_write_or(wal, @@ -774,21 +730,6 @@ static void kbl_gt_workarounds_init(struct drm_i915_private *i915) wa_write_or(wal, GEN9_GAMT_ECO_REG_RW_IA, GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); - - /* WaKBLVECSSemaphoreWaitPoll:kbl */ - if (IS_KBL_REVID(i915, KBL_REVID_A0, KBL_REVID_E0)) { - struct intel_engine_cs *engine; - unsigned int tmp; - - for_each_engine(engine, i915, tmp) { - if (engine->id == RCS) - continue; - - wa_write(wal, - RING_SEMA_WAIT_POLL(engine->mmio_base), - 1); - } - } } static void glk_gt_workarounds_init(struct drm_i915_private *i915) @@ -802,11 +743,6 @@ static void cfl_gt_workarounds_init(struct drm_i915_private *i915) gen9_gt_workarounds_init(i915); - /* WaEnableGapsTsvCreditFix:cfl */ - wa_write_or(wal, - GEN8_GARBCNTL, - GEN9_GAPS_TSV_CREDIT_DISABLE); - /* WaDisableGafsUnitClkGating:cfl */ wa_write_or(wal, GEN7_UCGCTL4, @@ -897,11 +833,6 @@ static void cnl_gt_workarounds_init(struct drm_i915_private *i915) wa_write_or(wal, GEN9_GAMT_ECO_REG_RW_IA, GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); - - /* WaEnablePreemptionGranularityControlByUMD:cnl */ - wa_masked_en(wal, - GEN7_FF_SLICE_CS_CHICKEN1, - GEN9_FFSC_PERCTX_PREEMPT_CTRL); } static void icl_gt_workarounds_init(struct drm_i915_private *i915) @@ -910,53 +841,17 @@ static void icl_gt_workarounds_init(struct drm_i915_private *i915) wa_init_mcr(i915); - /* This is not an Wa. Enable for better image quality */ - wa_masked_en(wal, - _3D_CHICKEN3, - _3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLE); - /* WaInPlaceDecompressionHang:icl */ wa_write_or(wal, GEN9_GAMT_ECO_REG_RW_IA, GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); - /* WaPipelineFlushCoherentLines:icl */ - wa_write_or(wal, - GEN8_L3SQCREG4, - GEN8_LQSC_FLUSH_COHERENT_LINES); - - /* Wa_1405543622:icl - * Formerly known as WaGAPZPriorityScheme - */ - wa_write_or(wal, - GEN8_GARBCNTL, - GEN11_ARBITRATION_PRIO_ORDER_MASK); - - /* Wa_1604223664:icl - * Formerly known as WaL3BankAddressHashing - */ - wa_write_masked_or(wal, - GEN8_GARBCNTL, - GEN11_HASH_CTRL_EXCL_MASK, - GEN11_HASH_CTRL_EXCL_BIT0); - wa_write_masked_or(wal, - GEN11_GLBLINVL, - GEN11_BANK_HASH_ADDR_EXCL_MASK, - GEN11_BANK_HASH_ADDR_EXCL_BIT0); - /* WaModifyGamTlbPartitioning:icl */ wa_write_masked_or(wal, GEN11_GACB_PERF_CTRL, GEN11_HASH_CTRL_MASK, GEN11_HASH_CTRL_BIT0 | GEN11_HASH_CTRL_BIT4); - /* Wa_1405733216:icl - * Formerly known as WaDisableCleanEvicts - */ - wa_write_or(wal, - GEN8_L3SQCREG4, - GEN11_LQSC_CLEAN_EVICT_DISABLE); - /* Wa_1405766107:icl * Formerly known as WaCL2SFHalfMaxAlloc */ @@ -989,26 +884,12 @@ static void icl_gt_workarounds_init(struct drm_i915_private *i915) INF_UNIT_LEVEL_CLKGATE, CGPSF_CLKGATE_DIS); - /* WaForwardProgressSoftReset:icl */ - wa_write_or(wal, - GEN10_SCRATCH_LNCF2, - PMFLUSHDONE_LNICRSDROP | - PMFLUSH_GAPL3UNBLOCK | - PMFLUSHDONE_LNEBLK); - /* Wa_1406463099:icl * Formerly known as WaGamTlbPendError */ wa_write_or(wal, GAMT_CHKN_BIT_REG, GAMT_CHKN_DISABLE_L3_COH_PIPE); - - /* Wa_1406609255:icl (pre-prod) */ - if (IS_ICL_REVID(i915, ICL_REVID_A0, ICL_REVID_B0)) - wa_write_or(wal, - GEN7_SARCHKMD, - GEN7_DISABLE_DEMAND_PREFETCH | - GEN7_DISABLE_SAMPLER_PREFETCH); } void intel_gt_init_workarounds(struct drm_i915_private *i915) @@ -1245,6 +1126,148 @@ void intel_whitelist_workarounds_apply(struct intel_engine_cs *engine) whitelist_apply(engine, whitelist_build(engine, &w)); } +static void rcs_engine_wa_init(struct intel_engine_cs *engine) +{ + struct drm_i915_private *i915 = engine->i915; + struct i915_wa_list *wal = &engine->wa_list; + + if (IS_ICELAKE(i915)) { + /* This is not an Wa. Enable for better image quality */ + wa_masked_en(wal, + _3D_CHICKEN3, + _3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLE); + + /* WaPipelineFlushCoherentLines:icl */ + wa_write_or(wal, + GEN8_L3SQCREG4, + GEN8_LQSC_FLUSH_COHERENT_LINES); + + /* + * Wa_1405543622:icl + * Formerly known as WaGAPZPriorityScheme + */ + wa_write_or(wal, + GEN8_GARBCNTL, + GEN11_ARBITRATION_PRIO_ORDER_MASK); + + /* + * Wa_1604223664:icl + * Formerly known as WaL3BankAddressHashing + */ + wa_write_masked_or(wal, + GEN8_GARBCNTL, + GEN11_HASH_CTRL_EXCL_MASK, + GEN11_HASH_CTRL_EXCL_BIT0); + wa_write_masked_or(wal, + GEN11_GLBLINVL, + GEN11_BANK_HASH_ADDR_EXCL_MASK, + GEN11_BANK_HASH_ADDR_EXCL_BIT0); + + /* + * Wa_1405733216:icl + * Formerly known as WaDisableCleanEvicts + */ + wa_write_or(wal, + GEN8_L3SQCREG4, + GEN11_LQSC_CLEAN_EVICT_DISABLE); + + /* WaForwardProgressSoftReset:icl */ + wa_write_or(wal, + GEN10_SCRATCH_LNCF2, + PMFLUSHDONE_LNICRSDROP | + PMFLUSH_GAPL3UNBLOCK | + PMFLUSHDONE_LNEBLK); + + /* Wa_1406609255:icl (pre-prod) */ + if (IS_ICL_REVID(i915, ICL_REVID_A0, ICL_REVID_B0)) + wa_write_or(wal, + GEN7_SARCHKMD, + GEN7_DISABLE_DEMAND_PREFETCH | + GEN7_DISABLE_SAMPLER_PREFETCH); + } + + if (IS_GEN9(i915) || IS_CANNONLAKE(i915)) { + /* WaEnablePreemptionGranularityControlByUMD:skl,bxt,kbl,cfl,cnl */ + wa_masked_en(wal, + GEN7_FF_SLICE_CS_CHICKEN1, + GEN9_FFSC_PERCTX_PREEMPT_CTRL); + } + + if (IS_SKYLAKE(i915) || IS_KABYLAKE(i915) || IS_COFFEELAKE(i915)) { + /* WaEnableGapsTsvCreditFix:skl,kbl,cfl */ + wa_write_or(wal, + GEN8_GARBCNTL, + GEN9_GAPS_TSV_CREDIT_DISABLE); + } + + if (IS_BROXTON(i915)) { + /* WaDisablePooledEuLoadBalancingFix:bxt */ + wa_masked_en(wal, + FF_SLICE_CS_CHICKEN2, + GEN9_POOLED_EU_LOAD_BALANCING_FIX_DISABLE); + } + + if (IS_GEN9(i915)) { + /* WaContextSwitchWithConcurrentTLBInvalidate:skl,bxt,kbl,glk,cfl */ + wa_masked_en(wal, + GEN9_CSFE_CHICKEN1_RCS, + GEN9_PREEMPT_GPGPU_SYNC_SWITCH_DISABLE); + + /* WaEnableLbsSlaRetryTimerDecrement:skl,bxt,kbl,glk,cfl */ + wa_write_or(wal, + BDW_SCRATCH1, + GEN9_LBS_SLA_RETRY_TIMER_DECREMENT_ENABLE); + + /* WaProgramL3SqcReg1DefaultForPerf:bxt,glk */ + if (IS_GEN9_LP(i915)) + wa_write_masked_or(wal, + GEN8_L3SQCREG1, + L3_PRIO_CREDITS_MASK, + L3_GENERAL_PRIO_CREDITS(62) | + L3_HIGH_PRIO_CREDITS(2)); + + /* WaOCLCoherentLineFlush:skl,bxt,kbl,cfl */ + wa_write_or(wal, + GEN8_L3SQCREG4, + GEN8_LQSC_FLUSH_COHERENT_LINES); + } +} + +static void xcs_engine_wa_init(struct intel_engine_cs *engine) +{ + struct drm_i915_private *i915 = engine->i915; + struct i915_wa_list *wal = &engine->wa_list; + + /* WaKBLVECSSemaphoreWaitPoll:kbl */ + if (IS_KBL_REVID(i915, KBL_REVID_A0, KBL_REVID_E0)) { + wa_write(wal, + RING_SEMA_WAIT_POLL(engine->mmio_base), + 1); + } +} + +void intel_engine_init_workarounds(struct intel_engine_cs *engine) +{ + struct i915_wa_list *wal = &engine->wa_list; + + if (GEM_WARN_ON(INTEL_GEN(engine->i915) < 8)) + return; + + wa_init_start(wal, engine->name); + + if (engine->id == RCS) + rcs_engine_wa_init(engine); + else + xcs_engine_wa_init(engine); + + wa_init_finish(wal); +} + +void intel_engine_apply_workarounds(struct intel_engine_cs *engine) +{ + wa_list_apply(engine->i915, &engine->wa_list); +} + #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST) #include "selftests/intel_workarounds.c" #endif diff --git a/drivers/gpu/drm/i915/intel_workarounds.h b/drivers/gpu/drm/i915/intel_workarounds.h index 263106600fdc..979695a53964 100644 --- a/drivers/gpu/drm/i915/intel_workarounds.h +++ b/drivers/gpu/drm/i915/intel_workarounds.h @@ -35,4 +35,7 @@ void intel_gt_apply_workarounds(struct drm_i915_private *dev_priv); void intel_whitelist_workarounds_apply(struct intel_engine_cs *engine); +void intel_engine_init_workarounds(struct intel_engine_cs *engine); +void intel_engine_apply_workarounds(struct intel_engine_cs *engine); + #endif From 094304beb4e1f4bfb2b59385a12d7074f4485e98 Mon Sep 17 00:00:00 2001 From: Tvrtko Ursulin Date: Mon, 3 Dec 2018 12:50:10 +0000 Subject: [PATCH 91/98] drm/i915: Verify GT workaround state after GPU init Since we now have all the GT workarounds in a table, by adding a simple shared helper function we can now verify that their values are still applied after some interesting events in the lifetime of the driver. Initially we only do this after GPU initialization. v2: Chris Wilson: * Simplify verification by realizing it's a simple xor and and. * Remove verification from engine reset path. * Return bool straight away from the verify API. v3: * API rename. (Chris Wilson) Signed-off-by: Tvrtko Ursulin Reviewed-by: Chris Wilson Link: https://patchwork.freedesktop.org/patch/msgid/20181203125014.3219-4-tvrtko.ursulin@linux.intel.com --- drivers/gpu/drm/i915/i915_drv.c | 1 + drivers/gpu/drm/i915/i915_gem.c | 3 +++ drivers/gpu/drm/i915/intel_workarounds.c | 34 ++++++++++++++++++++++++ drivers/gpu/drm/i915/intel_workarounds.h | 2 ++ 4 files changed, 40 insertions(+) diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c index 6f497cf31ffc..b310a897a4ad 100644 --- a/drivers/gpu/drm/i915/i915_drv.c +++ b/drivers/gpu/drm/i915/i915_drv.c @@ -53,6 +53,7 @@ #include "i915_vgpu.h" #include "intel_drv.h" #include "intel_uc.h" +#include "intel_workarounds.h" static struct drm_driver driver; diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index ad5a115e47b6..35ecfea4e903 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -5300,7 +5300,10 @@ int i915_gem_init_hw(struct drm_i915_private *dev_priv) I915_WRITE(MI_PREDICATE_RESULT_2, IS_HSW_GT3(dev_priv) ? LOWER_SLICE_ENABLED : LOWER_SLICE_DISABLED); + /* Apply the GT workarounds... */ intel_gt_apply_workarounds(dev_priv); + /* ...and determine whether they are sticking. */ + intel_gt_verify_workarounds(dev_priv, "init"); i915_gem_init_swizzling(dev_priv); diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c index 847988d15e8a..592226a0c711 100644 --- a/drivers/gpu/drm/i915/intel_workarounds.c +++ b/drivers/gpu/drm/i915/intel_workarounds.c @@ -977,6 +977,40 @@ void intel_gt_apply_workarounds(struct drm_i915_private *dev_priv) wa_list_apply(dev_priv, &dev_priv->gt_wa_list); } +static bool +wa_verify(const struct i915_wa *wa, u32 cur, const char *name, const char *from) +{ + if ((cur ^ wa->val) & wa->mask) { + DRM_ERROR("%s workaround lost on %s! (%x=%x/%x, expected %x, mask=%x)\n", + name, from, i915_mmio_reg_offset(wa->reg), cur, + cur & wa->mask, wa->val, wa->mask); + + return false; + } + + return true; +} + +static bool wa_list_verify(struct drm_i915_private *dev_priv, + const struct i915_wa_list *wal, + const char *from) +{ + struct i915_wa *wa; + unsigned int i; + bool ok = true; + + for (i = 0, wa = wal->list; i < wal->count; i++, wa++) + ok &= wa_verify(wa, I915_READ(wa->reg), wal->name, from); + + return ok; +} + +bool intel_gt_verify_workarounds(struct drm_i915_private *dev_priv, + const char *from) +{ + return wa_list_verify(dev_priv, &dev_priv->gt_wa_list, from); +} + struct whitelist { i915_reg_t reg[RING_MAX_NONPRIV_SLOTS]; unsigned int count; diff --git a/drivers/gpu/drm/i915/intel_workarounds.h b/drivers/gpu/drm/i915/intel_workarounds.h index 979695a53964..8822e6035f8d 100644 --- a/drivers/gpu/drm/i915/intel_workarounds.h +++ b/drivers/gpu/drm/i915/intel_workarounds.h @@ -32,6 +32,8 @@ int intel_ctx_workarounds_emit(struct i915_request *rq); void intel_gt_init_workarounds(struct drm_i915_private *dev_priv); void intel_gt_apply_workarounds(struct drm_i915_private *dev_priv); +bool intel_gt_verify_workarounds(struct drm_i915_private *dev_priv, + const char *from); void intel_whitelist_workarounds_apply(struct intel_engine_cs *engine); From 28d6ccce73bec983d5038f5cee1064957816cf44 Mon Sep 17 00:00:00 2001 From: Tvrtko Ursulin Date: Mon, 3 Dec 2018 12:50:11 +0000 Subject: [PATCH 92/98] drm/i915/selftests: Add tests for GT and engine workaround verification Two simple selftests which test that both GT and engine workarounds are not lost after either a full GPU reset, or after the per-engine ones. (Including checks that one engine reset is not affecting workarounds not belonging to itself.) v2: * Rebase for series refactoring. * Add spinner for actual engine reset! * Add idle reset test as well. (Chris Wilson) * Share existing global_reset_lock. (Chris Wilson) v3: * intel_engine_verify_workarounds can be static. * API rename. (Chris Wilson) * Move global reset lock out of the loop. (Chris Wilson) v4: * Add missing rpm puts. (Chris Wilson) Signed-off-by: Tvrtko Ursulin Reviewed-by: Chris Wilson Link: https://patchwork.freedesktop.org/patch/msgid/20181203125014.3219-5-tvrtko.ursulin@linux.intel.com --- drivers/gpu/drm/i915/Makefile | 1 + drivers/gpu/drm/i915/intel_workarounds.c | 6 + drivers/gpu/drm/i915/selftests/igt_reset.c | 44 ++++++ drivers/gpu/drm/i915/selftests/igt_reset.h | 15 ++ .../gpu/drm/i915/selftests/intel_hangcheck.c | 51 ++----- .../drm/i915/selftests/intel_workarounds.c | 144 +++++++++++++++++- 6 files changed, 214 insertions(+), 47 deletions(-) create mode 100644 drivers/gpu/drm/i915/selftests/igt_reset.c create mode 100644 drivers/gpu/drm/i915/selftests/igt_reset.h diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile index 50a8fa8fce64..19b5fe5016bf 100644 --- a/drivers/gpu/drm/i915/Makefile +++ b/drivers/gpu/drm/i915/Makefile @@ -166,6 +166,7 @@ i915-$(CONFIG_DRM_I915_SELFTEST) += \ selftests/i915_random.o \ selftests/i915_selftest.o \ selftests/igt_flush_test.o \ + selftests/igt_reset.o \ selftests/igt_spinner.o # virtual gpu code diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c index 592226a0c711..c53d4388930b 100644 --- a/drivers/gpu/drm/i915/intel_workarounds.c +++ b/drivers/gpu/drm/i915/intel_workarounds.c @@ -1303,5 +1303,11 @@ void intel_engine_apply_workarounds(struct intel_engine_cs *engine) } #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST) +static bool intel_engine_verify_workarounds(struct intel_engine_cs *engine, + const char *from) +{ + return wa_list_verify(engine->i915, &engine->wa_list, from); +} + #include "selftests/intel_workarounds.c" #endif diff --git a/drivers/gpu/drm/i915/selftests/igt_reset.c b/drivers/gpu/drm/i915/selftests/igt_reset.c new file mode 100644 index 000000000000..208a966da8ca --- /dev/null +++ b/drivers/gpu/drm/i915/selftests/igt_reset.c @@ -0,0 +1,44 @@ +/* + * SPDX-License-Identifier: MIT + * + * Copyright © 2018 Intel Corporation + */ + +#include "igt_reset.h" + +#include "../i915_drv.h" +#include "../intel_ringbuffer.h" + +void igt_global_reset_lock(struct drm_i915_private *i915) +{ + struct intel_engine_cs *engine; + enum intel_engine_id id; + + pr_debug("%s: current gpu_error=%08lx\n", + __func__, i915->gpu_error.flags); + + while (test_and_set_bit(I915_RESET_BACKOFF, &i915->gpu_error.flags)) + wait_event(i915->gpu_error.reset_queue, + !test_bit(I915_RESET_BACKOFF, + &i915->gpu_error.flags)); + + for_each_engine(engine, i915, id) { + while (test_and_set_bit(I915_RESET_ENGINE + id, + &i915->gpu_error.flags)) + wait_on_bit(&i915->gpu_error.flags, + I915_RESET_ENGINE + id, + TASK_UNINTERRUPTIBLE); + } +} + +void igt_global_reset_unlock(struct drm_i915_private *i915) +{ + struct intel_engine_cs *engine; + enum intel_engine_id id; + + for_each_engine(engine, i915, id) + clear_bit(I915_RESET_ENGINE + id, &i915->gpu_error.flags); + + clear_bit(I915_RESET_BACKOFF, &i915->gpu_error.flags); + wake_up_all(&i915->gpu_error.reset_queue); +} diff --git a/drivers/gpu/drm/i915/selftests/igt_reset.h b/drivers/gpu/drm/i915/selftests/igt_reset.h new file mode 100644 index 000000000000..5f0234d045d5 --- /dev/null +++ b/drivers/gpu/drm/i915/selftests/igt_reset.h @@ -0,0 +1,15 @@ +/* + * SPDX-License-Identifier: MIT + * + * Copyright © 2018 Intel Corporation + */ + +#ifndef __I915_SELFTESTS_IGT_RESET_H__ +#define __I915_SELFTESTS_IGT_RESET_H__ + +#include "../i915_drv.h" + +void igt_global_reset_lock(struct drm_i915_private *i915); +void igt_global_reset_unlock(struct drm_i915_private *i915); + +#endif diff --git a/drivers/gpu/drm/i915/selftests/intel_hangcheck.c b/drivers/gpu/drm/i915/selftests/intel_hangcheck.c index a48fbe2557ea..40efbed611de 100644 --- a/drivers/gpu/drm/i915/selftests/intel_hangcheck.c +++ b/drivers/gpu/drm/i915/selftests/intel_hangcheck.c @@ -27,6 +27,7 @@ #include "../i915_selftest.h" #include "i915_random.h" #include "igt_flush_test.h" +#include "igt_reset.h" #include "igt_wedge_me.h" #include "mock_context.h" @@ -354,40 +355,6 @@ unlock: return err; } -static void global_reset_lock(struct drm_i915_private *i915) -{ - struct intel_engine_cs *engine; - enum intel_engine_id id; - - pr_debug("%s: current gpu_error=%08lx\n", - __func__, i915->gpu_error.flags); - - while (test_and_set_bit(I915_RESET_BACKOFF, &i915->gpu_error.flags)) - wait_event(i915->gpu_error.reset_queue, - !test_bit(I915_RESET_BACKOFF, - &i915->gpu_error.flags)); - - for_each_engine(engine, i915, id) { - while (test_and_set_bit(I915_RESET_ENGINE + id, - &i915->gpu_error.flags)) - wait_on_bit(&i915->gpu_error.flags, - I915_RESET_ENGINE + id, - TASK_UNINTERRUPTIBLE); - } -} - -static void global_reset_unlock(struct drm_i915_private *i915) -{ - struct intel_engine_cs *engine; - enum intel_engine_id id; - - for_each_engine(engine, i915, id) - clear_bit(I915_RESET_ENGINE + id, &i915->gpu_error.flags); - - clear_bit(I915_RESET_BACKOFF, &i915->gpu_error.flags); - wake_up_all(&i915->gpu_error.reset_queue); -} - static int igt_global_reset(void *arg) { struct drm_i915_private *i915 = arg; @@ -396,7 +363,7 @@ static int igt_global_reset(void *arg) /* Check that we can issue a global GPU reset */ - global_reset_lock(i915); + igt_global_reset_lock(i915); set_bit(I915_RESET_HANDOFF, &i915->gpu_error.flags); mutex_lock(&i915->drm.struct_mutex); @@ -411,7 +378,7 @@ static int igt_global_reset(void *arg) mutex_unlock(&i915->drm.struct_mutex); GEM_BUG_ON(test_bit(I915_RESET_HANDOFF, &i915->gpu_error.flags)); - global_reset_unlock(i915); + igt_global_reset_unlock(i915); if (i915_terminally_wedged(&i915->gpu_error)) err = -EIO; @@ -942,7 +909,7 @@ static int igt_reset_wait(void *arg) /* Check that we detect a stuck waiter and issue a reset */ - global_reset_lock(i915); + igt_global_reset_lock(i915); mutex_lock(&i915->drm.struct_mutex); err = hang_init(&h, i915); @@ -994,7 +961,7 @@ fini: hang_fini(&h); unlock: mutex_unlock(&i915->drm.struct_mutex); - global_reset_unlock(i915); + igt_global_reset_unlock(i915); if (i915_terminally_wedged(&i915->gpu_error)) return -EIO; @@ -1072,7 +1039,7 @@ static int __igt_reset_evict_vma(struct drm_i915_private *i915, /* Check that we can recover an unbind stuck on a hanging request */ - global_reset_lock(i915); + igt_global_reset_lock(i915); mutex_lock(&i915->drm.struct_mutex); err = hang_init(&h, i915); @@ -1192,7 +1159,7 @@ fini: hang_fini(&h); unlock: mutex_unlock(&i915->drm.struct_mutex); - global_reset_unlock(i915); + igt_global_reset_unlock(i915); if (i915_terminally_wedged(&i915->gpu_error)) return -EIO; @@ -1272,7 +1239,7 @@ static int igt_reset_queue(void *arg) /* Check that we replay pending requests following a hang */ - global_reset_lock(i915); + igt_global_reset_lock(i915); mutex_lock(&i915->drm.struct_mutex); err = hang_init(&h, i915); @@ -1403,7 +1370,7 @@ fini: hang_fini(&h); unlock: mutex_unlock(&i915->drm.struct_mutex); - global_reset_unlock(i915); + igt_global_reset_unlock(i915); if (i915_terminally_wedged(&i915->gpu_error)) return -EIO; diff --git a/drivers/gpu/drm/i915/selftests/intel_workarounds.c b/drivers/gpu/drm/i915/selftests/intel_workarounds.c index 80396b3592f5..d76a048c3954 100644 --- a/drivers/gpu/drm/i915/selftests/intel_workarounds.c +++ b/drivers/gpu/drm/i915/selftests/intel_workarounds.c @@ -6,6 +6,8 @@ #include "../i915_selftest.h" +#include "igt_flush_test.h" +#include "igt_reset.h" #include "igt_spinner.h" #include "igt_wedge_me.h" #include "mock_context.h" @@ -290,7 +292,6 @@ static int live_reset_whitelist(void *arg) { struct drm_i915_private *i915 = arg; struct intel_engine_cs *engine = i915->engine[RCS]; - struct i915_gpu_error *error = &i915->gpu_error; struct whitelist w; int err = 0; @@ -302,8 +303,7 @@ static int live_reset_whitelist(void *arg) if (!whitelist_build(engine, &w)) return 0; - set_bit(I915_RESET_BACKOFF, &error->flags); - set_bit(I915_RESET_ENGINE + engine->id, &error->flags); + igt_global_reset_lock(i915); if (intel_has_reset_engine(i915)) { err = check_whitelist_across_reset(engine, @@ -322,15 +322,149 @@ static int live_reset_whitelist(void *arg) } out: - clear_bit(I915_RESET_ENGINE + engine->id, &error->flags); - clear_bit(I915_RESET_BACKOFF, &error->flags); + igt_global_reset_unlock(i915); return err; } +static bool verify_gt_engine_wa(struct drm_i915_private *i915, const char *str) +{ + struct intel_engine_cs *engine; + enum intel_engine_id id; + bool ok = true; + + ok &= intel_gt_verify_workarounds(i915, str); + + for_each_engine(engine, i915, id) + ok &= intel_engine_verify_workarounds(engine, str); + + return ok; +} + +static int +live_gpu_reset_gt_engine_workarounds(void *arg) +{ + struct drm_i915_private *i915 = arg; + struct i915_gpu_error *error = &i915->gpu_error; + bool ok; + + if (!intel_has_gpu_reset(i915)) + return 0; + + pr_info("Verifying after GPU reset...\n"); + + igt_global_reset_lock(i915); + + ok = verify_gt_engine_wa(i915, "before reset"); + if (!ok) + goto out; + + intel_runtime_pm_get(i915); + set_bit(I915_RESET_HANDOFF, &error->flags); + i915_reset(i915, ALL_ENGINES, "live_workarounds"); + intel_runtime_pm_put(i915); + + ok = verify_gt_engine_wa(i915, "after reset"); + +out: + igt_global_reset_unlock(i915); + + return ok ? 0 : -ESRCH; +} + +static int +live_engine_reset_gt_engine_workarounds(void *arg) +{ + struct drm_i915_private *i915 = arg; + struct intel_engine_cs *engine; + struct i915_gem_context *ctx; + struct igt_spinner spin; + enum intel_engine_id id; + struct i915_request *rq; + int ret = 0; + + if (!intel_has_reset_engine(i915)) + return 0; + + ctx = kernel_context(i915); + if (IS_ERR(ctx)) + return PTR_ERR(ctx); + + igt_global_reset_lock(i915); + + for_each_engine(engine, i915, id) { + bool ok; + + pr_info("Verifying after %s reset...\n", engine->name); + + ok = verify_gt_engine_wa(i915, "before reset"); + if (!ok) { + ret = -ESRCH; + goto err; + } + + intel_runtime_pm_get(i915); + i915_reset_engine(engine, "live_workarounds"); + intel_runtime_pm_put(i915); + + ok = verify_gt_engine_wa(i915, "after idle reset"); + if (!ok) { + ret = -ESRCH; + goto err; + } + + ret = igt_spinner_init(&spin, i915); + if (ret) + goto err; + + intel_runtime_pm_get(i915); + + rq = igt_spinner_create_request(&spin, ctx, engine, MI_NOOP); + if (IS_ERR(rq)) { + ret = PTR_ERR(rq); + igt_spinner_fini(&spin); + intel_runtime_pm_put(i915); + goto err; + } + + i915_request_add(rq); + + if (!igt_wait_for_spinner(&spin, rq)) { + pr_err("Spinner failed to start\n"); + igt_spinner_fini(&spin); + intel_runtime_pm_put(i915); + ret = -ETIMEDOUT; + goto err; + } + + i915_reset_engine(engine, "live_workarounds"); + + intel_runtime_pm_put(i915); + + igt_spinner_end(&spin); + igt_spinner_fini(&spin); + + ok = verify_gt_engine_wa(i915, "after busy reset"); + if (!ok) { + ret = -ESRCH; + goto err; + } + } + +err: + igt_global_reset_unlock(i915); + kernel_context_close(ctx); + + igt_flush_test(i915, I915_WAIT_LOCKED); + + return ret; +} + int intel_workarounds_live_selftests(struct drm_i915_private *i915) { static const struct i915_subtest tests[] = { SUBTEST(live_reset_whitelist), + SUBTEST(live_gpu_reset_gt_engine_workarounds), + SUBTEST(live_engine_reset_gt_engine_workarounds), }; int err; From 69bcdecf1af5600dabbab890e3d8d9714638f91d Mon Sep 17 00:00:00 2001 From: Tvrtko Ursulin Date: Mon, 3 Dec 2018 12:50:12 +0000 Subject: [PATCH 93/98] drm/i915: Move register white-listing to the common workaround framework Instead of having a separate list of white-listed registers we can trivially move this to the common workarounds framework. This brings us one step closer to the goal of driving all workaround classes using the same code. v2: * Use GEM_DEBUG_WARN_ON for the sanity check. (Chris Wilson) v3: * API rename. (Chris Wilson) Signed-off-by: Tvrtko Ursulin Reviewed-by: Chris Wilson Link: https://patchwork.freedesktop.org/patch/msgid/20181203125014.3219-6-tvrtko.ursulin@linux.intel.com --- drivers/gpu/drm/i915/intel_engine_cs.c | 1 + drivers/gpu/drm/i915/intel_lrc.c | 5 +- drivers/gpu/drm/i915/intel_ringbuffer.h | 1 + drivers/gpu/drm/i915/intel_workarounds.c | 83 ++++++++----------- drivers/gpu/drm/i915/intel_workarounds.h | 3 +- .../drm/i915/selftests/intel_workarounds.c | 40 ++++----- 6 files changed, 60 insertions(+), 73 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c index ef5d202e9d45..496462d77ebc 100644 --- a/drivers/gpu/drm/i915/intel_engine_cs.c +++ b/drivers/gpu/drm/i915/intel_engine_cs.c @@ -725,6 +725,7 @@ void intel_engine_cleanup_common(struct intel_engine_cs *engine) i915_timeline_fini(&engine->timeline); intel_wa_list_free(&engine->wa_list); + intel_wa_list_free(&engine->whitelist); } u64 intel_engine_get_active_head(const struct intel_engine_cs *engine) diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c index a5d6663ceafd..92e1f08e1483 100644 --- a/drivers/gpu/drm/i915/intel_lrc.c +++ b/drivers/gpu/drm/i915/intel_lrc.c @@ -1652,7 +1652,7 @@ static int gen8_init_render_ring(struct intel_engine_cs *engine) if (ret) return ret; - intel_whitelist_workarounds_apply(engine); + intel_engine_apply_whitelist(engine); /* We need to disable the AsyncFlip performance optimisations in order * to use MI_WAIT_FOR_EVENT within the CS. It should already be @@ -1675,7 +1675,7 @@ static int gen9_init_render_ring(struct intel_engine_cs *engine) if (ret) return ret; - intel_whitelist_workarounds_apply(engine); + intel_engine_apply_whitelist(engine); return 0; } @@ -2307,6 +2307,7 @@ int logical_render_ring_init(struct intel_engine_cs *engine) ret); } + intel_engine_init_whitelist(engine); intel_engine_init_workarounds(engine); return 0; diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h index 3abac391a739..7b110e221749 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.h +++ b/drivers/gpu/drm/i915/intel_ringbuffer.h @@ -437,6 +437,7 @@ struct intel_engine_cs { struct intel_hw_status_page status_page; struct i915_ctx_workarounds wa_ctx; struct i915_wa_list wa_list; + struct i915_wa_list whitelist; struct i915_vma *scratch; u32 irq_keep_mask; /* always keep these interrupts */ diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c index c53d4388930b..d920a6256c83 100644 --- a/drivers/gpu/drm/i915/intel_workarounds.c +++ b/drivers/gpu/drm/i915/intel_workarounds.c @@ -1011,29 +1011,20 @@ bool intel_gt_verify_workarounds(struct drm_i915_private *dev_priv, return wa_list_verify(dev_priv, &dev_priv->gt_wa_list, from); } -struct whitelist { - i915_reg_t reg[RING_MAX_NONPRIV_SLOTS]; - unsigned int count; - u32 nopid; -}; - -static void whitelist_reg(struct whitelist *w, i915_reg_t reg) +static void +whitelist_reg(struct i915_wa_list *wal, i915_reg_t reg) { - if (GEM_DEBUG_WARN_ON(w->count >= RING_MAX_NONPRIV_SLOTS)) + struct i915_wa wa = { + .reg = reg + }; + + if (GEM_DEBUG_WARN_ON(wal->count >= RING_MAX_NONPRIV_SLOTS)) return; - w->reg[w->count++] = reg; + wal_add(wal, &wa); } -static void bdw_whitelist_build(struct whitelist *w) -{ -} - -static void chv_whitelist_build(struct whitelist *w) -{ -} - -static void gen9_whitelist_build(struct whitelist *w) +static void gen9_whitelist_build(struct i915_wa_list *w) { /* WaVFEStateAfterPipeControlwithMediaStateClear:skl,bxt,glk,cfl */ whitelist_reg(w, GEN9_CTX_PREEMPT_REG); @@ -1045,7 +1036,7 @@ static void gen9_whitelist_build(struct whitelist *w) whitelist_reg(w, GEN8_HDC_CHICKEN1); } -static void skl_whitelist_build(struct whitelist *w) +static void skl_whitelist_build(struct i915_wa_list *w) { gen9_whitelist_build(w); @@ -1053,12 +1044,12 @@ static void skl_whitelist_build(struct whitelist *w) whitelist_reg(w, GEN8_L3SQCREG4); } -static void bxt_whitelist_build(struct whitelist *w) +static void bxt_whitelist_build(struct i915_wa_list *w) { gen9_whitelist_build(w); } -static void kbl_whitelist_build(struct whitelist *w) +static void kbl_whitelist_build(struct i915_wa_list *w) { gen9_whitelist_build(w); @@ -1066,7 +1057,7 @@ static void kbl_whitelist_build(struct whitelist *w) whitelist_reg(w, GEN8_L3SQCREG4); } -static void glk_whitelist_build(struct whitelist *w) +static void glk_whitelist_build(struct i915_wa_list *w) { gen9_whitelist_build(w); @@ -1074,18 +1065,18 @@ static void glk_whitelist_build(struct whitelist *w) whitelist_reg(w, GEN9_SLICE_COMMON_ECO_CHICKEN1); } -static void cfl_whitelist_build(struct whitelist *w) +static void cfl_whitelist_build(struct i915_wa_list *w) { gen9_whitelist_build(w); } -static void cnl_whitelist_build(struct whitelist *w) +static void cnl_whitelist_build(struct i915_wa_list *w) { /* WaEnablePreemptionGranularityControlByUMD:cnl */ whitelist_reg(w, GEN8_CS_CHICKEN1); } -static void icl_whitelist_build(struct whitelist *w) +static void icl_whitelist_build(struct i915_wa_list *w) { /* WaAllowUMDToModifyHalfSliceChicken7:icl */ whitelist_reg(w, GEN9_HALF_SLICE_CHICKEN7); @@ -1094,22 +1085,21 @@ static void icl_whitelist_build(struct whitelist *w) whitelist_reg(w, GEN10_SAMPLER_MODE); } -static struct whitelist *whitelist_build(struct intel_engine_cs *engine, - struct whitelist *w) +void intel_engine_init_whitelist(struct intel_engine_cs *engine) { struct drm_i915_private *i915 = engine->i915; + struct i915_wa_list *w = &engine->whitelist; GEM_BUG_ON(engine->id != RCS); - w->count = 0; - w->nopid = i915_mmio_reg_offset(RING_NOPID(engine->mmio_base)); + wa_init_start(w, "whitelist"); if (INTEL_GEN(i915) < 8) - return NULL; + return; else if (IS_BROADWELL(i915)) - bdw_whitelist_build(w); + return; else if (IS_CHERRYVIEW(i915)) - chv_whitelist_build(w); + return; else if (IS_SKYLAKE(i915)) skl_whitelist_build(w); else if (IS_BROXTON(i915)) @@ -1127,37 +1117,30 @@ static struct whitelist *whitelist_build(struct intel_engine_cs *engine, else MISSING_CASE(INTEL_GEN(i915)); - return w; + wa_init_finish(w); } -static void whitelist_apply(struct intel_engine_cs *engine, - const struct whitelist *w) +void intel_engine_apply_whitelist(struct intel_engine_cs *engine) { struct drm_i915_private *dev_priv = engine->i915; + const struct i915_wa_list *wal = &engine->whitelist; const u32 base = engine->mmio_base; + struct i915_wa *wa; unsigned int i; - if (!w) + if (!wal->count) return; - intel_uncore_forcewake_get(engine->i915, FORCEWAKE_ALL); - - for (i = 0; i < w->count; i++) - I915_WRITE_FW(RING_FORCE_TO_NONPRIV(base, i), - i915_mmio_reg_offset(w->reg[i])); + for (i = 0, wa = wal->list; i < wal->count; i++, wa++) + I915_WRITE(RING_FORCE_TO_NONPRIV(base, i), + i915_mmio_reg_offset(wa->reg)); /* And clear the rest just in case of garbage */ for (; i < RING_MAX_NONPRIV_SLOTS; i++) - I915_WRITE_FW(RING_FORCE_TO_NONPRIV(base, i), w->nopid); + I915_WRITE(RING_FORCE_TO_NONPRIV(base, i), + i915_mmio_reg_offset(RING_NOPID(base))); - intel_uncore_forcewake_put(engine->i915, FORCEWAKE_ALL); -} - -void intel_whitelist_workarounds_apply(struct intel_engine_cs *engine) -{ - struct whitelist w; - - whitelist_apply(engine, whitelist_build(engine, &w)); + DRM_DEBUG_DRIVER("Applied %u %s workarounds\n", wal->count, wal->name); } static void rcs_engine_wa_init(struct intel_engine_cs *engine) diff --git a/drivers/gpu/drm/i915/intel_workarounds.h b/drivers/gpu/drm/i915/intel_workarounds.h index 8822e6035f8d..3f99bfcb4a03 100644 --- a/drivers/gpu/drm/i915/intel_workarounds.h +++ b/drivers/gpu/drm/i915/intel_workarounds.h @@ -35,7 +35,8 @@ void intel_gt_apply_workarounds(struct drm_i915_private *dev_priv); bool intel_gt_verify_workarounds(struct drm_i915_private *dev_priv, const char *from); -void intel_whitelist_workarounds_apply(struct intel_engine_cs *engine); +void intel_engine_init_whitelist(struct intel_engine_cs *engine); +void intel_engine_apply_whitelist(struct intel_engine_cs *engine); void intel_engine_init_workarounds(struct intel_engine_cs *engine); void intel_engine_apply_workarounds(struct intel_engine_cs *engine); diff --git a/drivers/gpu/drm/i915/selftests/intel_workarounds.c b/drivers/gpu/drm/i915/selftests/intel_workarounds.c index d76a048c3954..67017d5175b8 100644 --- a/drivers/gpu/drm/i915/selftests/intel_workarounds.c +++ b/drivers/gpu/drm/i915/selftests/intel_workarounds.c @@ -94,17 +94,23 @@ err_obj: return ERR_PTR(err); } -static u32 get_whitelist_reg(const struct whitelist *w, unsigned int i) +static u32 +get_whitelist_reg(const struct intel_engine_cs *engine, unsigned int i) { - return i < w->count ? i915_mmio_reg_offset(w->reg[i]) : w->nopid; + i915_reg_t reg = i < engine->whitelist.count ? + engine->whitelist.list[i].reg : + RING_NOPID(engine->mmio_base); + + return i915_mmio_reg_offset(reg); } -static void print_results(const struct whitelist *w, const u32 *results) +static void +print_results(const struct intel_engine_cs *engine, const u32 *results) { unsigned int i; for (i = 0; i < RING_MAX_NONPRIV_SLOTS; i++) { - u32 expected = get_whitelist_reg(w, i); + u32 expected = get_whitelist_reg(engine, i); u32 actual = results[i]; pr_info("RING_NONPRIV[%d]: expected 0x%08x, found 0x%08x\n", @@ -112,8 +118,7 @@ static void print_results(const struct whitelist *w, const u32 *results) } } -static int check_whitelist(const struct whitelist *w, - struct i915_gem_context *ctx, +static int check_whitelist(struct i915_gem_context *ctx, struct intel_engine_cs *engine) { struct drm_i915_gem_object *results; @@ -141,11 +146,11 @@ static int check_whitelist(const struct whitelist *w, } for (i = 0; i < RING_MAX_NONPRIV_SLOTS; i++) { - u32 expected = get_whitelist_reg(w, i); + u32 expected = get_whitelist_reg(engine, i); u32 actual = vaddr[i]; if (expected != actual) { - print_results(w, vaddr); + print_results(engine, vaddr); pr_err("Invalid RING_NONPRIV[%d], expected 0x%08x, found 0x%08x\n", i, expected, actual); @@ -217,7 +222,6 @@ err: static int check_whitelist_across_reset(struct intel_engine_cs *engine, int (*reset)(struct intel_engine_cs *), - const struct whitelist *w, const char *name) { struct drm_i915_private *i915 = engine->i915; @@ -227,7 +231,7 @@ static int check_whitelist_across_reset(struct intel_engine_cs *engine, int err; pr_info("Checking %d whitelisted registers (RING_NONPRIV) [%s]\n", - w->count, name); + engine->whitelist.count, name); if (want_spin) { err = igt_spinner_init(&spin, i915); @@ -239,7 +243,7 @@ static int check_whitelist_across_reset(struct intel_engine_cs *engine, if (IS_ERR(ctx)) return PTR_ERR(ctx); - err = check_whitelist(w, ctx, engine); + err = check_whitelist(ctx, engine); if (err) { pr_err("Invalid whitelist *before* %s reset!\n", name); goto out; @@ -263,7 +267,7 @@ static int check_whitelist_across_reset(struct intel_engine_cs *engine, goto out; } - err = check_whitelist(w, ctx, engine); + err = check_whitelist(ctx, engine); if (err) { pr_err("Whitelist not preserved in context across %s reset!\n", name); @@ -276,7 +280,7 @@ static int check_whitelist_across_reset(struct intel_engine_cs *engine, if (IS_ERR(ctx)) return PTR_ERR(ctx); - err = check_whitelist(w, ctx, engine); + err = check_whitelist(ctx, engine); if (err) { pr_err("Invalid whitelist *after* %s reset in fresh context!\n", name); @@ -292,22 +296,18 @@ static int live_reset_whitelist(void *arg) { struct drm_i915_private *i915 = arg; struct intel_engine_cs *engine = i915->engine[RCS]; - struct whitelist w; int err = 0; /* If we reset the gpu, we should not lose the RING_NONPRIV */ - if (!engine) - return 0; - - if (!whitelist_build(engine, &w)) + if (!engine || engine->whitelist.count == 0) return 0; igt_global_reset_lock(i915); if (intel_has_reset_engine(i915)) { err = check_whitelist_across_reset(engine, - do_engine_reset, &w, + do_engine_reset, "engine"); if (err) goto out; @@ -315,7 +315,7 @@ static int live_reset_whitelist(void *arg) if (intel_has_gpu_reset(i915)) { err = check_whitelist_across_reset(engine, - do_device_reset, &w, + do_device_reset, "device"); if (err) goto out; From 452420d22d5b41256a0bb82402a797295e525da9 Mon Sep 17 00:00:00 2001 From: Tvrtko Ursulin Date: Mon, 3 Dec 2018 13:33:57 +0000 Subject: [PATCH 94/98] drm/i915: Fuse per-context workaround handling with the common framework Convert the per context workaround handling code to run against the newly introduced common workaround framework and fuse the two to use the existing smarter list add helper, the one which does the sorted insert and merges registers where possible. This completes migration of all four classes of workarounds onto the common framework. Existing macros are kept untouched for smaller code churn. v2: * Rename to list name ctx_wa_list and move from dev_priv to engine. v3: * API rename and parameters tweaking. (Chris Wilson) Signed-off-by: Tvrtko Ursulin Reviewed-by: Chris Wilson Link: https://patchwork.freedesktop.org/patch/msgid/20181203133357.10341-1-tvrtko.ursulin@linux.intel.com --- drivers/gpu/drm/i915/i915_debugfs.c | 12 +- drivers/gpu/drm/i915/i915_drv.h | 15 -- drivers/gpu/drm/i915/i915_gem_context.c | 6 +- drivers/gpu/drm/i915/intel_engine_cs.c | 1 + drivers/gpu/drm/i915/intel_lrc.c | 2 +- drivers/gpu/drm/i915/intel_ringbuffer.c | 2 +- drivers/gpu/drm/i915/intel_ringbuffer.h | 1 + drivers/gpu/drm/i915/intel_workarounds.c | 329 +++++++++++------------ drivers/gpu/drm/i915/intel_workarounds.h | 5 +- 9 files changed, 166 insertions(+), 207 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c index 129b9a6f8309..38dcee1ca062 100644 --- a/drivers/gpu/drm/i915/i915_debugfs.c +++ b/drivers/gpu/drm/i915/i915_debugfs.c @@ -3375,13 +3375,15 @@ static int i915_shared_dplls_info(struct seq_file *m, void *unused) static int i915_wa_registers(struct seq_file *m, void *unused) { - struct i915_workarounds *wa = &node_to_i915(m->private)->workarounds; - int i; + struct drm_i915_private *i915 = node_to_i915(m->private); + const struct i915_wa_list *wal = &i915->engine[RCS]->ctx_wa_list; + struct i915_wa *wa; + unsigned int i; - seq_printf(m, "Workarounds applied: %d\n", wa->count); - for (i = 0; i < wa->count; ++i) + seq_printf(m, "Workarounds applied: %u\n", wal->count); + for (i = 0, wa = wal->list; i < wal->count; i++, wa++) seq_printf(m, "0x%X: 0x%08X, mask: 0x%08X\n", - wa->reg[i].addr, wa->reg[i].value, wa->reg[i].mask); + i915_mmio_reg_offset(wa->reg), wa->val, wa->mask); return 0; } diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index d725390d5a48..23a3dc6f3907 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -1190,20 +1190,6 @@ struct i915_frontbuffer_tracking { unsigned flip_bits; }; -struct i915_wa_reg { - u32 addr; - u32 value; - /* bitmask representing WA bits */ - u32 mask; -}; - -#define I915_MAX_WA_REGS 16 - -struct i915_workarounds { - struct i915_wa_reg reg[I915_MAX_WA_REGS]; - u32 count; -}; - struct i915_virtual_gpu { bool active; u32 caps; @@ -1653,7 +1639,6 @@ struct drm_i915_private { int dpio_phy_iosf_port[I915_NUM_PHYS_VLV]; - struct i915_workarounds workarounds; struct i915_wa_list gt_wa_list; struct i915_frontbuffer_tracking fb_tracking; diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c index b97963db0287..371c07087095 100644 --- a/drivers/gpu/drm/i915/i915_gem_context.c +++ b/drivers/gpu/drm/i915/i915_gem_context.c @@ -535,16 +535,12 @@ static bool needs_preempt_context(struct drm_i915_private *i915) int i915_gem_contexts_init(struct drm_i915_private *dev_priv) { struct i915_gem_context *ctx; - int ret; /* Reassure ourselves we are only called once */ GEM_BUG_ON(dev_priv->kernel_context); GEM_BUG_ON(dev_priv->preempt_context); - ret = intel_ctx_workarounds_init(dev_priv); - if (ret) - return ret; - + intel_engine_init_ctx_wa(dev_priv->engine[RCS]); init_contexts(dev_priv); /* lowest priority; idle task */ diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c index 496462d77ebc..6b427bc52f78 100644 --- a/drivers/gpu/drm/i915/intel_engine_cs.c +++ b/drivers/gpu/drm/i915/intel_engine_cs.c @@ -724,6 +724,7 @@ void intel_engine_cleanup_common(struct intel_engine_cs *engine) i915_timeline_fini(&engine->timeline); + intel_wa_list_free(&engine->ctx_wa_list); intel_wa_list_free(&engine->wa_list); intel_wa_list_free(&engine->whitelist); } diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c index 92e1f08e1483..87227fd9ae5f 100644 --- a/drivers/gpu/drm/i915/intel_lrc.c +++ b/drivers/gpu/drm/i915/intel_lrc.c @@ -2087,7 +2087,7 @@ static int gen8_init_rcs_context(struct i915_request *rq) { int ret; - ret = intel_ctx_workarounds_emit(rq); + ret = intel_engine_emit_ctx_wa(rq); if (ret) return ret; diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index 81b10d85b738..7f88df5bff09 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -656,7 +656,7 @@ static int intel_rcs_ctx_init(struct i915_request *rq) { int ret; - ret = intel_ctx_workarounds_emit(rq); + ret = intel_engine_emit_ctx_wa(rq); if (ret != 0) return ret; diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h index 7b110e221749..927bb21a2b0b 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.h +++ b/drivers/gpu/drm/i915/intel_ringbuffer.h @@ -436,6 +436,7 @@ struct intel_engine_cs { struct intel_hw_status_page status_page; struct i915_ctx_workarounds wa_ctx; + struct i915_wa_list ctx_wa_list; struct i915_wa_list wa_list; struct i915_wa_list whitelist; struct i915_vma *scratch; diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c index d920a6256c83..e52bd3f5d526 100644 --- a/drivers/gpu/drm/i915/intel_workarounds.c +++ b/drivers/gpu/drm/i915/intel_workarounds.c @@ -59,61 +59,87 @@ static void wa_init_finish(struct i915_wa_list *wal) return; DRM_DEBUG_DRIVER("Initialized %u %s workarounds\n", - wal->count, wal->name); + wal->wa_count, wal->name); } -static void wa_add(struct drm_i915_private *i915, - i915_reg_t reg, const u32 mask, const u32 val) +static void _wa_add(struct i915_wa_list *wal, const struct i915_wa *wa) { - struct i915_workarounds *wa = &i915->workarounds; - unsigned int start = 0, end = wa->count; - unsigned int addr = i915_mmio_reg_offset(reg); - struct i915_wa_reg *r; + unsigned int addr = i915_mmio_reg_offset(wa->reg); + unsigned int start = 0, end = wal->count; + const unsigned int grow = 1 << 4; + struct i915_wa *wa_; + + GEM_BUG_ON(!is_power_of_2(grow)); + + if (IS_ALIGNED(wal->count, grow)) { /* Either uninitialized or full. */ + struct i915_wa *list; + + list = kmalloc_array(ALIGN(wal->count + 1, grow), sizeof(*wa), + GFP_KERNEL); + if (!list) { + DRM_ERROR("No space for workaround init!\n"); + return; + } + + if (wal->list) + memcpy(list, wal->list, sizeof(*wa) * wal->count); + + wal->list = list; + } while (start < end) { unsigned int mid = start + (end - start) / 2; - if (wa->reg[mid].addr < addr) { + if (i915_mmio_reg_offset(wal->list[mid].reg) < addr) { start = mid + 1; - } else if (wa->reg[mid].addr > addr) { + } else if (i915_mmio_reg_offset(wal->list[mid].reg) > addr) { end = mid; } else { - r = &wa->reg[mid]; + wa_ = &wal->list[mid]; - if ((mask & ~r->mask) == 0) { + if ((wa->mask & ~wa_->mask) == 0) { DRM_ERROR("Discarding overwritten w/a for reg %04x (mask: %08x, value: %08x)\n", - addr, r->mask, r->value); + i915_mmio_reg_offset(wa_->reg), + wa_->mask, wa_->val); - r->value &= ~mask; + wa_->val &= ~wa->mask; } - r->value |= val; - r->mask |= mask; + wal->wa_count++; + wa_->val |= wa->val; + wa_->mask |= wa->mask; return; } } - if (WARN_ON_ONCE(wa->count >= I915_MAX_WA_REGS)) { - DRM_ERROR("Dropping w/a for reg %04x (mask: %08x, value: %08x)\n", - addr, mask, val); - return; - } + wal->wa_count++; + wa_ = &wal->list[wal->count++]; + *wa_ = *wa; - r = &wa->reg[wa->count++]; - r->addr = addr; - r->value = val; - r->mask = mask; - - while (r-- > wa->reg) { - GEM_BUG_ON(r[0].addr == r[1].addr); - if (r[1].addr > r[0].addr) + while (wa_-- > wal->list) { + GEM_BUG_ON(i915_mmio_reg_offset(wa_[0].reg) == + i915_mmio_reg_offset(wa_[1].reg)); + if (i915_mmio_reg_offset(wa_[1].reg) > + i915_mmio_reg_offset(wa_[0].reg)) break; - swap(r[1], r[0]); + swap(wa_[1], wa_[0]); } } -#define WA_REG(addr, mask, val) wa_add(dev_priv, (addr), (mask), (val)) +static void +__wa_add(struct i915_wa_list *wal, i915_reg_t reg, u32 mask, u32 val) +{ + struct i915_wa wa = { + .reg = reg, + .mask = mask, + .val = val + }; + + _wa_add(wal, &wa); +} + +#define WA_REG(addr, mask, val) __wa_add(wal, (addr), (mask), (val)) #define WA_SET_BIT_MASKED(addr, mask) \ WA_REG(addr, (mask), _MASKED_BIT_ENABLE(mask)) @@ -124,8 +150,10 @@ static void wa_add(struct drm_i915_private *i915, #define WA_SET_FIELD_MASKED(addr, mask, value) \ WA_REG(addr, (mask), _MASKED_FIELD(mask, value)) -static int gen8_ctx_workarounds_init(struct drm_i915_private *dev_priv) +static void gen8_ctx_workarounds_init(struct intel_engine_cs *engine) { + struct i915_wa_list *wal = &engine->ctx_wa_list; + WA_SET_BIT_MASKED(INSTPM, INSTPM_FORCE_ORDERING); /* WaDisableAsyncFlipPerfMode:bdw,chv */ @@ -169,17 +197,14 @@ static int gen8_ctx_workarounds_init(struct drm_i915_private *dev_priv) WA_SET_FIELD_MASKED(GEN7_GT_MODE, GEN6_WIZ_HASHING_MASK, GEN6_WIZ_HASHING_16x4); - - return 0; } -static int bdw_ctx_workarounds_init(struct drm_i915_private *dev_priv) +static void bdw_ctx_workarounds_init(struct intel_engine_cs *engine) { - int ret; + struct drm_i915_private *i915 = engine->i915; + struct i915_wa_list *wal = &engine->ctx_wa_list; - ret = gen8_ctx_workarounds_init(dev_priv); - if (ret) - return ret; + gen8_ctx_workarounds_init(engine); /* WaDisableThreadStallDopClockGating:bdw (pre-production) */ WA_SET_BIT_MASKED(GEN8_ROW_CHICKEN, STALL_DOP_GATING_DISABLE); @@ -199,31 +224,28 @@ static int bdw_ctx_workarounds_init(struct drm_i915_private *dev_priv) /* WaForceContextSaveRestoreNonCoherent:bdw */ HDC_FORCE_CONTEXT_SAVE_RESTORE_NON_COHERENT | /* WaDisableFenceDestinationToSLM:bdw (pre-prod) */ - (IS_BDW_GT3(dev_priv) ? HDC_FENCE_DEST_SLM_DISABLE : 0)); - - return 0; + (IS_BDW_GT3(i915) ? HDC_FENCE_DEST_SLM_DISABLE : 0)); } -static int chv_ctx_workarounds_init(struct drm_i915_private *dev_priv) +static void chv_ctx_workarounds_init(struct intel_engine_cs *engine) { - int ret; + struct i915_wa_list *wal = &engine->ctx_wa_list; - ret = gen8_ctx_workarounds_init(dev_priv); - if (ret) - return ret; + gen8_ctx_workarounds_init(engine); /* WaDisableThreadStallDopClockGating:chv */ WA_SET_BIT_MASKED(GEN8_ROW_CHICKEN, STALL_DOP_GATING_DISABLE); /* Improve HiZ throughput on CHV. */ WA_SET_BIT_MASKED(HIZ_CHICKEN, CHV_HZ_8X8_MODE_IN_1X); - - return 0; } -static int gen9_ctx_workarounds_init(struct drm_i915_private *dev_priv) +static void gen9_ctx_workarounds_init(struct intel_engine_cs *engine) { - if (HAS_LLC(dev_priv)) { + struct drm_i915_private *i915 = engine->i915; + struct i915_wa_list *wal = &engine->ctx_wa_list; + + if (HAS_LLC(i915)) { /* WaCompressedResourceSamplerPbeMediaNewHashMode:skl,kbl * * Must match Display Engine. See @@ -242,7 +264,7 @@ static int gen9_ctx_workarounds_init(struct drm_i915_private *dev_priv) PARTIAL_INSTRUCTION_SHOOTDOWN_DISABLE); /* Syncing dependencies between camera and graphics:skl,bxt,kbl */ - if (!IS_COFFEELAKE(dev_priv)) + if (!IS_COFFEELAKE(i915)) WA_SET_BIT_MASKED(HALF_SLICE_CHICKEN3, GEN9_DISABLE_OCL_OOB_SUPPRESS_LOGIC); @@ -285,9 +307,7 @@ static int gen9_ctx_workarounds_init(struct drm_i915_private *dev_priv) HDC_FORCE_NON_COHERENT); /* WaDisableSamplerPowerBypassForSOPingPong:skl,bxt,kbl,cfl */ - if (IS_SKYLAKE(dev_priv) || - IS_KABYLAKE(dev_priv) || - IS_COFFEELAKE(dev_priv)) + if (IS_SKYLAKE(i915) || IS_KABYLAKE(i915) || IS_COFFEELAKE(i915)) WA_SET_BIT_MASKED(HALF_SLICE_CHICKEN3, GEN8_SAMPLER_POWER_BYPASS_DIS); @@ -314,14 +334,14 @@ static int gen9_ctx_workarounds_init(struct drm_i915_private *dev_priv) GEN9_PREEMPT_GPGPU_COMMAND_LEVEL); /* WaClearHIZ_WM_CHICKEN3:bxt,glk */ - if (IS_GEN9_LP(dev_priv)) + if (IS_GEN9_LP(i915)) WA_SET_BIT_MASKED(GEN9_WM_CHICKEN3, GEN9_FACTOR_IN_CLR_VAL_HIZ); - - return 0; } -static int skl_tune_iz_hashing(struct drm_i915_private *dev_priv) +static void skl_tune_iz_hashing(struct intel_engine_cs *engine) { + struct drm_i915_private *i915 = engine->i915; + struct i915_wa_list *wal = &engine->ctx_wa_list; u8 vals[3] = { 0, 0, 0 }; unsigned int i; @@ -332,7 +352,7 @@ static int skl_tune_iz_hashing(struct drm_i915_private *dev_priv) * Only consider slices where one, and only one, subslice has 7 * EUs */ - if (!is_power_of_2(INTEL_INFO(dev_priv)->sseu.subslice_7eu[i])) + if (!is_power_of_2(INTEL_INFO(i915)->sseu.subslice_7eu[i])) continue; /* @@ -341,12 +361,12 @@ static int skl_tune_iz_hashing(struct drm_i915_private *dev_priv) * * -> 0 <= ss <= 3; */ - ss = ffs(INTEL_INFO(dev_priv)->sseu.subslice_7eu[i]) - 1; + ss = ffs(INTEL_INFO(i915)->sseu.subslice_7eu[i]) - 1; vals[i] = 3 - ss; } if (vals[0] == 0 && vals[1] == 0 && vals[2] == 0) - return 0; + return; /* Tune IZ hashing. See intel_device_info_runtime_init() */ WA_SET_FIELD_MASKED(GEN7_GT_MODE, @@ -356,28 +376,19 @@ static int skl_tune_iz_hashing(struct drm_i915_private *dev_priv) GEN9_IZ_HASHING(2, vals[2]) | GEN9_IZ_HASHING(1, vals[1]) | GEN9_IZ_HASHING(0, vals[0])); - - return 0; } -static int skl_ctx_workarounds_init(struct drm_i915_private *dev_priv) +static void skl_ctx_workarounds_init(struct intel_engine_cs *engine) { - int ret; - - ret = gen9_ctx_workarounds_init(dev_priv); - if (ret) - return ret; - - return skl_tune_iz_hashing(dev_priv); + gen9_ctx_workarounds_init(engine); + skl_tune_iz_hashing(engine); } -static int bxt_ctx_workarounds_init(struct drm_i915_private *dev_priv) +static void bxt_ctx_workarounds_init(struct intel_engine_cs *engine) { - int ret; + struct i915_wa_list *wal = &engine->ctx_wa_list; - ret = gen9_ctx_workarounds_init(dev_priv); - if (ret) - return ret; + gen9_ctx_workarounds_init(engine); /* WaDisableThreadStallDopClockGating:bxt */ WA_SET_BIT_MASKED(GEN8_ROW_CHICKEN, @@ -386,52 +397,41 @@ static int bxt_ctx_workarounds_init(struct drm_i915_private *dev_priv) /* WaToEnableHwFixForPushConstHWBug:bxt */ WA_SET_BIT_MASKED(COMMON_SLICE_CHICKEN2, GEN8_SBE_DISABLE_REPLAY_BUF_OPTIMIZATION); - - return 0; } -static int kbl_ctx_workarounds_init(struct drm_i915_private *dev_priv) +static void kbl_ctx_workarounds_init(struct intel_engine_cs *engine) { - int ret; + struct drm_i915_private *i915 = engine->i915; + struct i915_wa_list *wal = &engine->ctx_wa_list; - ret = gen9_ctx_workarounds_init(dev_priv); - if (ret) - return ret; + gen9_ctx_workarounds_init(engine); /* WaToEnableHwFixForPushConstHWBug:kbl */ - if (IS_KBL_REVID(dev_priv, KBL_REVID_C0, REVID_FOREVER)) + if (IS_KBL_REVID(i915, KBL_REVID_C0, REVID_FOREVER)) WA_SET_BIT_MASKED(COMMON_SLICE_CHICKEN2, GEN8_SBE_DISABLE_REPLAY_BUF_OPTIMIZATION); /* WaDisableSbeCacheDispatchPortSharing:kbl */ WA_SET_BIT_MASKED(GEN7_HALF_SLICE_CHICKEN1, GEN7_SBE_SS_CACHE_DISPATCH_PORT_SHARING_DISABLE); - - return 0; } -static int glk_ctx_workarounds_init(struct drm_i915_private *dev_priv) +static void glk_ctx_workarounds_init(struct intel_engine_cs *engine) { - int ret; + struct i915_wa_list *wal = &engine->ctx_wa_list; - ret = gen9_ctx_workarounds_init(dev_priv); - if (ret) - return ret; + gen9_ctx_workarounds_init(engine); /* WaToEnableHwFixForPushConstHWBug:glk */ WA_SET_BIT_MASKED(COMMON_SLICE_CHICKEN2, GEN8_SBE_DISABLE_REPLAY_BUF_OPTIMIZATION); - - return 0; } -static int cfl_ctx_workarounds_init(struct drm_i915_private *dev_priv) +static void cfl_ctx_workarounds_init(struct intel_engine_cs *engine) { - int ret; + struct i915_wa_list *wal = &engine->ctx_wa_list; - ret = gen9_ctx_workarounds_init(dev_priv); - if (ret) - return ret; + gen9_ctx_workarounds_init(engine); /* WaToEnableHwFixForPushConstHWBug:cfl */ WA_SET_BIT_MASKED(COMMON_SLICE_CHICKEN2, @@ -440,18 +440,19 @@ static int cfl_ctx_workarounds_init(struct drm_i915_private *dev_priv) /* WaDisableSbeCacheDispatchPortSharing:cfl */ WA_SET_BIT_MASKED(GEN7_HALF_SLICE_CHICKEN1, GEN7_SBE_SS_CACHE_DISPATCH_PORT_SHARING_DISABLE); - - return 0; } -static int cnl_ctx_workarounds_init(struct drm_i915_private *dev_priv) +static void cnl_ctx_workarounds_init(struct intel_engine_cs *engine) { + struct drm_i915_private *i915 = engine->i915; + struct i915_wa_list *wal = &engine->ctx_wa_list; + /* WaForceContextSaveRestoreNonCoherent:cnl */ WA_SET_BIT_MASKED(CNL_HDC_CHICKEN0, HDC_FORCE_CONTEXT_SAVE_RESTORE_NON_COHERENT); /* WaThrottleEUPerfToAvoidTDBackPressure:cnl(pre-prod) */ - if (IS_CNL_REVID(dev_priv, CNL_REVID_B0, CNL_REVID_B0)) + if (IS_CNL_REVID(i915, CNL_REVID_B0, CNL_REVID_B0)) WA_SET_BIT_MASKED(GEN8_ROW_CHICKEN, THROTTLE_12_5); /* WaDisableReplayBufferBankArbitrationOptimization:cnl */ @@ -459,7 +460,7 @@ static int cnl_ctx_workarounds_init(struct drm_i915_private *dev_priv) GEN8_SBE_DISABLE_REPLAY_BUF_OPTIMIZATION); /* WaDisableEnhancedSBEVertexCaching:cnl (pre-prod) */ - if (IS_CNL_REVID(dev_priv, 0, CNL_REVID_B0)) + if (IS_CNL_REVID(i915, 0, CNL_REVID_B0)) WA_SET_BIT_MASKED(COMMON_SLICE_CHICKEN2, GEN8_CSC2_SBE_VUE_CACHE_CONSERVATIVE); @@ -479,16 +480,17 @@ static int cnl_ctx_workarounds_init(struct drm_i915_private *dev_priv) /* WaDisableEarlyEOT:cnl */ WA_SET_BIT_MASKED(GEN8_ROW_CHICKEN, DISABLE_EARLY_EOT); - - return 0; } -static int icl_ctx_workarounds_init(struct drm_i915_private *dev_priv) +static void icl_ctx_workarounds_init(struct intel_engine_cs *engine) { + struct drm_i915_private *i915 = engine->i915; + struct i915_wa_list *wal = &engine->ctx_wa_list; + /* Wa_1604370585:icl (pre-prod) * Formerly known as WaPushConstantDereferenceHoldDisable */ - if (IS_ICL_REVID(dev_priv, ICL_REVID_A0, ICL_REVID_B0)) + if (IS_ICL_REVID(i915, ICL_REVID_A0, ICL_REVID_B0)) WA_SET_BIT_MASKED(GEN7_ROW_CHICKEN2, PUSH_CONSTANT_DEREF_DISABLE); @@ -504,7 +506,7 @@ static int icl_ctx_workarounds_init(struct drm_i915_private *dev_priv) /* Wa_2006611047:icl (pre-prod) * Formerly known as WaDisableImprovedTdlClkGating */ - if (IS_ICL_REVID(dev_priv, ICL_REVID_A0, ICL_REVID_A0)) + if (IS_ICL_REVID(i915, ICL_REVID_A0, ICL_REVID_A0)) WA_SET_BIT_MASKED(GEN7_ROW_CHICKEN2, GEN11_TDL_CLOCK_GATING_FIX_DISABLE); @@ -513,70 +515,67 @@ static int icl_ctx_workarounds_init(struct drm_i915_private *dev_priv) GEN11_STATE_CACHE_REDIRECT_TO_CS); /* Wa_2006665173:icl (pre-prod) */ - if (IS_ICL_REVID(dev_priv, ICL_REVID_A0, ICL_REVID_A0)) + if (IS_ICL_REVID(i915, ICL_REVID_A0, ICL_REVID_A0)) WA_SET_BIT_MASKED(GEN11_COMMON_SLICE_CHICKEN3, GEN11_BLEND_EMB_FIX_DISABLE_IN_RCC); - - return 0; } -int intel_ctx_workarounds_init(struct drm_i915_private *dev_priv) +void intel_engine_init_ctx_wa(struct intel_engine_cs *engine) { - int err = 0; + struct drm_i915_private *i915 = engine->i915; + struct i915_wa_list *wal = &engine->ctx_wa_list; - dev_priv->workarounds.count = 0; + wa_init_start(wal, "context"); - if (INTEL_GEN(dev_priv) < 8) - err = 0; - else if (IS_BROADWELL(dev_priv)) - err = bdw_ctx_workarounds_init(dev_priv); - else if (IS_CHERRYVIEW(dev_priv)) - err = chv_ctx_workarounds_init(dev_priv); - else if (IS_SKYLAKE(dev_priv)) - err = skl_ctx_workarounds_init(dev_priv); - else if (IS_BROXTON(dev_priv)) - err = bxt_ctx_workarounds_init(dev_priv); - else if (IS_KABYLAKE(dev_priv)) - err = kbl_ctx_workarounds_init(dev_priv); - else if (IS_GEMINILAKE(dev_priv)) - err = glk_ctx_workarounds_init(dev_priv); - else if (IS_COFFEELAKE(dev_priv)) - err = cfl_ctx_workarounds_init(dev_priv); - else if (IS_CANNONLAKE(dev_priv)) - err = cnl_ctx_workarounds_init(dev_priv); - else if (IS_ICELAKE(dev_priv)) - err = icl_ctx_workarounds_init(dev_priv); + if (INTEL_GEN(i915) < 8) + return; + else if (IS_BROADWELL(i915)) + bdw_ctx_workarounds_init(engine); + else if (IS_CHERRYVIEW(i915)) + chv_ctx_workarounds_init(engine); + else if (IS_SKYLAKE(i915)) + skl_ctx_workarounds_init(engine); + else if (IS_BROXTON(i915)) + bxt_ctx_workarounds_init(engine); + else if (IS_KABYLAKE(i915)) + kbl_ctx_workarounds_init(engine); + else if (IS_GEMINILAKE(i915)) + glk_ctx_workarounds_init(engine); + else if (IS_COFFEELAKE(i915)) + cfl_ctx_workarounds_init(engine); + else if (IS_CANNONLAKE(i915)) + cnl_ctx_workarounds_init(engine); + else if (IS_ICELAKE(i915)) + icl_ctx_workarounds_init(engine); else - MISSING_CASE(INTEL_GEN(dev_priv)); - if (err) - return err; + MISSING_CASE(INTEL_GEN(i915)); - DRM_DEBUG_DRIVER("Number of context specific w/a: %d\n", - dev_priv->workarounds.count); - return 0; + wa_init_finish(wal); } -int intel_ctx_workarounds_emit(struct i915_request *rq) +int intel_engine_emit_ctx_wa(struct i915_request *rq) { - struct i915_workarounds *w = &rq->i915->workarounds; + struct i915_wa_list *wal = &rq->engine->ctx_wa_list; + struct i915_wa *wa; + unsigned int i; u32 *cs; - int ret, i; + int ret; - if (w->count == 0) + if (wal->count == 0) return 0; ret = rq->engine->emit_flush(rq, EMIT_BARRIER); if (ret) return ret; - cs = intel_ring_begin(rq, (w->count * 2 + 2)); + cs = intel_ring_begin(rq, (wal->count * 2 + 2)); if (IS_ERR(cs)) return PTR_ERR(cs); - *cs++ = MI_LOAD_REGISTER_IMM(w->count); - for (i = 0; i < w->count; i++) { - *cs++ = w->reg[i].addr; - *cs++ = w->reg[i].value; + *cs++ = MI_LOAD_REGISTER_IMM(wal->count); + for (i = 0, wa = wal->list; i < wal->count; i++, wa++) { + *cs++ = i915_mmio_reg_offset(wa->reg); + *cs++ = wa->val; } *cs++ = MI_NOOP; @@ -589,32 +588,6 @@ int intel_ctx_workarounds_emit(struct i915_request *rq) return 0; } -static void -wal_add(struct i915_wa_list *wal, const struct i915_wa *wa) -{ - const unsigned int grow = 1 << 4; - - GEM_BUG_ON(!is_power_of_2(grow)); - - if (IS_ALIGNED(wal->count, grow)) { /* Either uninitialized or full. */ - struct i915_wa *list; - - list = kmalloc_array(ALIGN(wal->count + 1, grow), sizeof(*wa), - GFP_KERNEL); - if (!list) { - DRM_ERROR("No space for workaround init!\n"); - return; - } - - if (wal->list) - memcpy(list, wal->list, sizeof(*wa) * wal->count); - - wal->list = list; - } - - wal->list[wal->count++] = *wa; -} - static void wa_masked_en(struct i915_wa_list *wal, i915_reg_t reg, u32 val) { @@ -624,7 +597,7 @@ wa_masked_en(struct i915_wa_list *wal, i915_reg_t reg, u32 val) .val = _MASKED_BIT_ENABLE(val) }; - wal_add(wal, &wa); + _wa_add(wal, &wa); } static void @@ -637,7 +610,7 @@ wa_write_masked_or(struct i915_wa_list *wal, i915_reg_t reg, u32 mask, .val = val }; - wal_add(wal, &wa); + _wa_add(wal, &wa); } static void @@ -1021,7 +994,7 @@ whitelist_reg(struct i915_wa_list *wal, i915_reg_t reg) if (GEM_DEBUG_WARN_ON(wal->count >= RING_MAX_NONPRIV_SLOTS)) return; - wal_add(wal, &wa); + _wa_add(wal, &wa); } static void gen9_whitelist_build(struct i915_wa_list *w) diff --git a/drivers/gpu/drm/i915/intel_workarounds.h b/drivers/gpu/drm/i915/intel_workarounds.h index 3f99bfcb4a03..7c734714b05e 100644 --- a/drivers/gpu/drm/i915/intel_workarounds.h +++ b/drivers/gpu/drm/i915/intel_workarounds.h @@ -19,6 +19,7 @@ struct i915_wa_list { const char *name; struct i915_wa *list; unsigned int count; + unsigned int wa_count; }; static inline void intel_wa_list_free(struct i915_wa_list *wal) @@ -27,8 +28,8 @@ static inline void intel_wa_list_free(struct i915_wa_list *wal) memset(wal, 0, sizeof(*wal)); } -int intel_ctx_workarounds_init(struct drm_i915_private *dev_priv); -int intel_ctx_workarounds_emit(struct i915_request *rq); +void intel_engine_init_ctx_wa(struct intel_engine_cs *engine); +int intel_engine_emit_ctx_wa(struct i915_request *rq); void intel_gt_init_workarounds(struct drm_i915_private *dev_priv); void intel_gt_apply_workarounds(struct drm_i915_private *dev_priv); From 4d8d9fc7050106cbac8141bd5ed5db3e4abbd5fa Mon Sep 17 00:00:00 2001 From: Tvrtko Ursulin Date: Mon, 3 Dec 2018 12:50:14 +0000 Subject: [PATCH 95/98] drm/i915: Trim unused workaround list entries The new workaround list allocator grows the list in chunks so will end up with some unused space. Trim it when the initialization phase is done to free up a tiny bit of slab. v2: * Simplify with kmemdup. (Chris Wilson) v3: * Refactor for __size removal. Signed-off-by: Tvrtko Ursulin Reviewed-by: Chris Wilson Link: https://patchwork.freedesktop.org/patch/msgid/20181203125014.3219-8-tvrtko.ursulin@linux.intel.com --- drivers/gpu/drm/i915/intel_workarounds.c | 16 +++++++++++++++- 1 file changed, 15 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c index e52bd3f5d526..4f41e326f3f3 100644 --- a/drivers/gpu/drm/i915/intel_workarounds.c +++ b/drivers/gpu/drm/i915/intel_workarounds.c @@ -53,8 +53,22 @@ static void wa_init_start(struct i915_wa_list *wal, const char *name) wal->name = name; } +#define WA_LIST_CHUNK (1 << 4) + static void wa_init_finish(struct i915_wa_list *wal) { + /* Trim unused entries. */ + if (!IS_ALIGNED(wal->count, WA_LIST_CHUNK)) { + struct i915_wa *list = kmemdup(wal->list, + wal->count * sizeof(*list), + GFP_KERNEL); + + if (list) { + kfree(wal->list); + wal->list = list; + } + } + if (!wal->count) return; @@ -66,7 +80,7 @@ static void _wa_add(struct i915_wa_list *wal, const struct i915_wa *wa) { unsigned int addr = i915_mmio_reg_offset(wa->reg); unsigned int start = 0, end = wal->count; - const unsigned int grow = 1 << 4; + const unsigned int grow = WA_LIST_CHUNK; struct i915_wa *wa_; GEM_BUG_ON(!is_power_of_2(grow)); From 5179749925933575a67f9d8f16d0cc204f98a29f Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Tue, 4 Dec 2018 14:15:16 +0000 Subject: [PATCH 96/98] drm/i915: Allocate a common scratch page Currently we allocate a scratch page for each engine, but since we only ever write into it for post-sync operations, it is not exposed to userspace nor do we care for coherency. As we then do not care about its contents, we can use one page for all, reducing our allocations and avoid complications by not assuming per-engine isolation. For later use, it simplifies engine initialisation (by removing the allocation that required struct_mutex!) and means that we can always rely on there being a scratch page. v2: Check that we allocated a large enough scratch for I830 w/a Fixes: 06e562e7f515 ("drm/i915/ringbuffer: Delay after EMIT_INVALIDATE for gen4/gen5") # v4.18.20 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=108850 Signed-off-by: Chris Wilson Cc: Tvrtko Ursulin Cc: Mika Kuoppala Reviewed-by: Mika Kuoppala Link: https://patchwork.freedesktop.org/patch/msgid/20181204141522.13640-1-chris@chris-wilson.co.uk Cc: Joonas Lahtinen Cc: # v4.18.20+ --- drivers/gpu/drm/i915/i915_drv.h | 7 ++++ drivers/gpu/drm/i915/i915_gem.c | 50 ++++++++++++++++++++++++- drivers/gpu/drm/i915/i915_gpu_error.c | 2 +- drivers/gpu/drm/i915/intel_engine_cs.c | 42 --------------------- drivers/gpu/drm/i915/intel_lrc.c | 17 +++------ drivers/gpu/drm/i915/intel_ringbuffer.c | 37 ++++++------------ drivers/gpu/drm/i915/intel_ringbuffer.h | 5 --- 7 files changed, 74 insertions(+), 86 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 23a3dc6f3907..c5f01964f0fb 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -1983,6 +1983,8 @@ struct drm_i915_private { struct delayed_work idle_work; ktime_t last_init_time; + + struct i915_vma *scratch; } gt; /* perform PHY state sanity checks? */ @@ -3713,4 +3715,9 @@ static inline int intel_hws_csb_write_index(struct drm_i915_private *i915) return I915_HWS_CSB_WRITE_INDEX; } +static inline u32 i915_scratch_offset(const struct drm_i915_private *i915) +{ + return i915_ggtt_offset(i915->gt.scratch); +} + #endif diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 35ecfea4e903..d36a9755ad91 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -5498,6 +5498,44 @@ err_active: goto out_ctx; } +static int +i915_gem_init_scratch(struct drm_i915_private *i915, unsigned int size) +{ + struct drm_i915_gem_object *obj; + struct i915_vma *vma; + int ret; + + obj = i915_gem_object_create_stolen(i915, size); + if (!obj) + obj = i915_gem_object_create_internal(i915, size); + if (IS_ERR(obj)) { + DRM_ERROR("Failed to allocate scratch page\n"); + return PTR_ERR(obj); + } + + vma = i915_vma_instance(obj, &i915->ggtt.vm, NULL); + if (IS_ERR(vma)) { + ret = PTR_ERR(vma); + goto err_unref; + } + + ret = i915_vma_pin(vma, 0, 0, PIN_GLOBAL | PIN_HIGH); + if (ret) + goto err_unref; + + i915->gt.scratch = vma; + return 0; + +err_unref: + i915_gem_object_put(obj); + return ret; +} + +static void i915_gem_fini_scratch(struct drm_i915_private *i915) +{ + i915_vma_unpin_and_release(&i915->gt.scratch, 0); +} + int i915_gem_init(struct drm_i915_private *dev_priv) { int ret; @@ -5544,12 +5582,19 @@ int i915_gem_init(struct drm_i915_private *dev_priv) goto err_unlock; } - ret = i915_gem_contexts_init(dev_priv); + ret = i915_gem_init_scratch(dev_priv, + IS_GEN2(dev_priv) ? SZ_256K : PAGE_SIZE); if (ret) { GEM_BUG_ON(ret == -EIO); goto err_ggtt; } + ret = i915_gem_contexts_init(dev_priv); + if (ret) { + GEM_BUG_ON(ret == -EIO); + goto err_scratch; + } + ret = intel_engines_init(dev_priv); if (ret) { GEM_BUG_ON(ret == -EIO); @@ -5622,6 +5667,8 @@ err_pm: err_context: if (ret != -EIO) i915_gem_contexts_fini(dev_priv); +err_scratch: + i915_gem_fini_scratch(dev_priv); err_ggtt: err_unlock: intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL); @@ -5673,6 +5720,7 @@ void i915_gem_fini(struct drm_i915_private *dev_priv) intel_uc_fini(dev_priv); i915_gem_cleanup_engines(dev_priv); i915_gem_contexts_fini(dev_priv); + i915_gem_fini_scratch(dev_priv); mutex_unlock(&dev_priv->drm.struct_mutex); intel_wa_list_free(&dev_priv->gt_wa_list); diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c index a6885a59568b..07465123c166 100644 --- a/drivers/gpu/drm/i915/i915_gpu_error.c +++ b/drivers/gpu/drm/i915/i915_gpu_error.c @@ -1571,7 +1571,7 @@ static void gem_record_rings(struct i915_gpu_state *error) if (HAS_BROKEN_CS_TLB(i915)) ee->wa_batchbuffer = i915_error_object_create(i915, - engine->scratch); + i915->gt.scratch); request_record_user_bo(request, ee); ee->ctx = diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c index 6b427bc52f78..af2873403009 100644 --- a/drivers/gpu/drm/i915/intel_engine_cs.c +++ b/drivers/gpu/drm/i915/intel_engine_cs.c @@ -493,46 +493,6 @@ void intel_engine_setup_common(struct intel_engine_cs *engine) intel_engine_init_cmd_parser(engine); } -int intel_engine_create_scratch(struct intel_engine_cs *engine, - unsigned int size) -{ - struct drm_i915_gem_object *obj; - struct i915_vma *vma; - int ret; - - WARN_ON(engine->scratch); - - obj = i915_gem_object_create_stolen(engine->i915, size); - if (!obj) - obj = i915_gem_object_create_internal(engine->i915, size); - if (IS_ERR(obj)) { - DRM_ERROR("Failed to allocate scratch page\n"); - return PTR_ERR(obj); - } - - vma = i915_vma_instance(obj, &engine->i915->ggtt.vm, NULL); - if (IS_ERR(vma)) { - ret = PTR_ERR(vma); - goto err_unref; - } - - ret = i915_vma_pin(vma, 0, 0, PIN_GLOBAL | PIN_HIGH); - if (ret) - goto err_unref; - - engine->scratch = vma; - return 0; - -err_unref: - i915_gem_object_put(obj); - return ret; -} - -void intel_engine_cleanup_scratch(struct intel_engine_cs *engine) -{ - i915_vma_unpin_and_release(&engine->scratch, 0); -} - static void cleanup_status_page(struct intel_engine_cs *engine) { if (HWS_NEEDS_PHYSICAL(engine->i915)) { @@ -707,8 +667,6 @@ void intel_engine_cleanup_common(struct intel_engine_cs *engine) { struct drm_i915_private *i915 = engine->i915; - intel_engine_cleanup_scratch(engine); - cleanup_status_page(engine); intel_engine_fini_breadcrumbs(engine); diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c index 87227fd9ae5f..d7fa301b5ec7 100644 --- a/drivers/gpu/drm/i915/intel_lrc.c +++ b/drivers/gpu/drm/i915/intel_lrc.c @@ -1288,9 +1288,10 @@ static int execlists_request_alloc(struct i915_request *request) static u32 * gen8_emit_flush_coherentl3_wa(struct intel_engine_cs *engine, u32 *batch) { + /* NB no one else is allowed to scribble over scratch + 256! */ *batch++ = MI_STORE_REGISTER_MEM_GEN8 | MI_SRM_LRM_GLOBAL_GTT; *batch++ = i915_mmio_reg_offset(GEN8_L3SQCREG4); - *batch++ = i915_ggtt_offset(engine->scratch) + 256; + *batch++ = i915_scratch_offset(engine->i915) + 256; *batch++ = 0; *batch++ = MI_LOAD_REGISTER_IMM(1); @@ -1304,7 +1305,7 @@ gen8_emit_flush_coherentl3_wa(struct intel_engine_cs *engine, u32 *batch) *batch++ = MI_LOAD_REGISTER_MEM_GEN8 | MI_SRM_LRM_GLOBAL_GTT; *batch++ = i915_mmio_reg_offset(GEN8_L3SQCREG4); - *batch++ = i915_ggtt_offset(engine->scratch) + 256; + *batch++ = i915_scratch_offset(engine->i915) + 256; *batch++ = 0; return batch; @@ -1341,7 +1342,7 @@ static u32 *gen8_init_indirectctx_bb(struct intel_engine_cs *engine, u32 *batch) PIPE_CONTROL_GLOBAL_GTT_IVB | PIPE_CONTROL_CS_STALL | PIPE_CONTROL_QW_WRITE, - i915_ggtt_offset(engine->scratch) + + i915_scratch_offset(engine->i915) + 2 * CACHELINE_BYTES); *batch++ = MI_ARB_ON_OFF | MI_ARB_ENABLE; @@ -1973,7 +1974,7 @@ static int gen8_emit_flush_render(struct i915_request *request, { struct intel_engine_cs *engine = request->engine; u32 scratch_addr = - i915_ggtt_offset(engine->scratch) + 2 * CACHELINE_BYTES; + i915_scratch_offset(engine->i915) + 2 * CACHELINE_BYTES; bool vf_flush_wa = false, dc_flush_wa = false; u32 *cs, flags = 0; int len; @@ -2292,10 +2293,6 @@ int logical_render_ring_init(struct intel_engine_cs *engine) if (ret) return ret; - ret = intel_engine_create_scratch(engine, PAGE_SIZE); - if (ret) - goto err_cleanup_common; - ret = intel_init_workaround_bb(engine); if (ret) { /* @@ -2311,10 +2308,6 @@ int logical_render_ring_init(struct intel_engine_cs *engine) intel_engine_init_workarounds(engine); return 0; - -err_cleanup_common: - intel_engine_cleanup_common(engine); - return ret; } int logical_xcs_ring_init(struct intel_engine_cs *engine) diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index 7f88df5bff09..c5eb26a7ee79 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -150,8 +150,7 @@ gen4_render_ring_flush(struct i915_request *rq, u32 mode) */ if (mode & EMIT_INVALIDATE) { *cs++ = GFX_OP_PIPE_CONTROL(4) | PIPE_CONTROL_QW_WRITE; - *cs++ = i915_ggtt_offset(rq->engine->scratch) | - PIPE_CONTROL_GLOBAL_GTT; + *cs++ = i915_scratch_offset(rq->i915) | PIPE_CONTROL_GLOBAL_GTT; *cs++ = 0; *cs++ = 0; @@ -159,8 +158,7 @@ gen4_render_ring_flush(struct i915_request *rq, u32 mode) *cs++ = MI_FLUSH; *cs++ = GFX_OP_PIPE_CONTROL(4) | PIPE_CONTROL_QW_WRITE; - *cs++ = i915_ggtt_offset(rq->engine->scratch) | - PIPE_CONTROL_GLOBAL_GTT; + *cs++ = i915_scratch_offset(rq->i915) | PIPE_CONTROL_GLOBAL_GTT; *cs++ = 0; *cs++ = 0; } @@ -212,8 +210,7 @@ gen4_render_ring_flush(struct i915_request *rq, u32 mode) static int intel_emit_post_sync_nonzero_flush(struct i915_request *rq) { - u32 scratch_addr = - i915_ggtt_offset(rq->engine->scratch) + 2 * CACHELINE_BYTES; + u32 scratch_addr = i915_scratch_offset(rq->i915) + 2 * CACHELINE_BYTES; u32 *cs; cs = intel_ring_begin(rq, 6); @@ -246,8 +243,7 @@ intel_emit_post_sync_nonzero_flush(struct i915_request *rq) static int gen6_render_ring_flush(struct i915_request *rq, u32 mode) { - u32 scratch_addr = - i915_ggtt_offset(rq->engine->scratch) + 2 * CACHELINE_BYTES; + u32 scratch_addr = i915_scratch_offset(rq->i915) + 2 * CACHELINE_BYTES; u32 *cs, flags = 0; int ret; @@ -316,8 +312,7 @@ gen7_render_ring_cs_stall_wa(struct i915_request *rq) static int gen7_render_ring_flush(struct i915_request *rq, u32 mode) { - u32 scratch_addr = - i915_ggtt_offset(rq->engine->scratch) + 2 * CACHELINE_BYTES; + u32 scratch_addr = i915_scratch_offset(rq->i915) + 2 * CACHELINE_BYTES; u32 *cs, flags = 0; /* @@ -994,7 +989,7 @@ i965_emit_bb_start(struct i915_request *rq, } /* Just userspace ABI convention to limit the wa batch bo to a resonable size */ -#define I830_BATCH_LIMIT (256*1024) +#define I830_BATCH_LIMIT SZ_256K #define I830_TLB_ENTRIES (2) #define I830_WA_SIZE max(I830_TLB_ENTRIES*4096, I830_BATCH_LIMIT) static int @@ -1002,7 +997,9 @@ i830_emit_bb_start(struct i915_request *rq, u64 offset, u32 len, unsigned int dispatch_flags) { - u32 *cs, cs_offset = i915_ggtt_offset(rq->engine->scratch); + u32 *cs, cs_offset = i915_scratch_offset(rq->i915); + + GEM_BUG_ON(rq->i915->gt.scratch->size < I830_WA_SIZE); cs = intel_ring_begin(rq, 6); if (IS_ERR(cs)) @@ -1459,7 +1456,6 @@ static int intel_init_ring_buffer(struct intel_engine_cs *engine) { struct i915_timeline *timeline; struct intel_ring *ring; - unsigned int size; int err; intel_engine_setup_common(engine); @@ -1484,21 +1480,12 @@ static int intel_init_ring_buffer(struct intel_engine_cs *engine) GEM_BUG_ON(engine->buffer); engine->buffer = ring; - size = PAGE_SIZE; - if (HAS_BROKEN_CS_TLB(engine->i915)) - size = I830_WA_SIZE; - err = intel_engine_create_scratch(engine, size); + err = intel_engine_init_common(engine); if (err) goto err_unpin; - err = intel_engine_init_common(engine); - if (err) - goto err_scratch; - return 0; -err_scratch: - intel_engine_cleanup_scratch(engine); err_unpin: intel_ring_unpin(ring); err_ring: @@ -1572,7 +1559,7 @@ static int flush_pd_dir(struct i915_request *rq) /* Stall until the page table load is complete */ *cs++ = MI_STORE_REGISTER_MEM | MI_SRM_LRM_GLOBAL_GTT; *cs++ = i915_mmio_reg_offset(RING_PP_DIR_BASE(engine)); - *cs++ = i915_ggtt_offset(engine->scratch); + *cs++ = i915_scratch_offset(rq->i915); *cs++ = MI_NOOP; intel_ring_advance(rq, cs); @@ -1681,7 +1668,7 @@ static inline int mi_set_context(struct i915_request *rq, u32 flags) /* Insert a delay before the next switch! */ *cs++ = MI_STORE_REGISTER_MEM | MI_SRM_LRM_GLOBAL_GTT; *cs++ = i915_mmio_reg_offset(last_reg); - *cs++ = i915_ggtt_offset(engine->scratch); + *cs++ = i915_scratch_offset(rq->i915); *cs++ = MI_NOOP; } *cs++ = MI_ARB_ON_OFF | MI_ARB_ENABLE; diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h index 927bb21a2b0b..72edaa7ff411 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.h +++ b/drivers/gpu/drm/i915/intel_ringbuffer.h @@ -439,7 +439,6 @@ struct intel_engine_cs { struct i915_wa_list ctx_wa_list; struct i915_wa_list wa_list; struct i915_wa_list whitelist; - struct i915_vma *scratch; u32 irq_keep_mask; /* always keep these interrupts */ u32 irq_enable_mask; /* bitmask to enable ring interrupt */ @@ -896,10 +895,6 @@ void intel_engine_setup_common(struct intel_engine_cs *engine); int intel_engine_init_common(struct intel_engine_cs *engine); void intel_engine_cleanup_common(struct intel_engine_cs *engine); -int intel_engine_create_scratch(struct intel_engine_cs *engine, - unsigned int size); -void intel_engine_cleanup_scratch(struct intel_engine_cs *engine); - int intel_init_render_ring_buffer(struct intel_engine_cs *engine); int intel_init_bsd_ring_buffer(struct intel_engine_cs *engine); int intel_init_blt_ring_buffer(struct intel_engine_cs *engine); From 0716931a82b4d0e211d2ef66616ad7130107e455 Mon Sep 17 00:00:00 2001 From: Jani Nikula Date: Tue, 4 Dec 2018 12:19:26 +0200 Subject: [PATCH 97/98] drm/i915/icl: fix transcoder state readout Commit 2ca711caeca2 ("drm/i915/icl: Consider DSI for getting transcoder state") clobbers the previously read TRANS_DDI_FUNC_CTL_EDP register contents with TRANS_DDI_FUNC_CTL_DSI0 contents. Fix the state readout, and handle DSI 1 while at it. Use a bitmask for iterating and logging transcoders, because the allowed combinations are a bit funky. Fixes: 2ca711caeca2 ("drm/i915/icl: Consider DSI for getting transcoder state") Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=108928 Cc: Ville Syrjala Cc: Madhav Chauhan Reviewed-by: Imre Deak Signed-off-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/20181204101926.17174-1-jani.nikula@intel.com --- drivers/gpu/drm/i915/intel_display.c | 56 ++++++++++++++++------------ 1 file changed, 33 insertions(+), 23 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index a2584f977ab1..07c861884c70 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -9476,13 +9476,18 @@ static bool hsw_get_transcoder_state(struct intel_crtc *crtc, struct drm_device *dev = crtc->base.dev; struct drm_i915_private *dev_priv = to_i915(dev); enum intel_display_power_domain power_domain; + unsigned long panel_transcoder_mask = BIT(TRANSCODER_EDP); + unsigned long enabled_panel_transcoders = 0; + enum transcoder panel_transcoder; u32 tmp; - bool is_dsi = false; - bool is_edp = false; + + if (IS_ICELAKE(dev_priv)) + panel_transcoder_mask |= + BIT(TRANSCODER_DSI_0) | BIT(TRANSCODER_DSI_1); /* * The pipe->transcoder mapping is fixed with the exception of the eDP - * transcoder handled below. + * and DSI transcoders handled below. */ pipe_config->cpu_transcoder = (enum transcoder) crtc->pipe; @@ -9490,23 +9495,26 @@ static bool hsw_get_transcoder_state(struct intel_crtc *crtc, * XXX: Do intel_display_power_get_if_enabled before reading this (for * consistency and less surprising code; it's in always on power). */ - tmp = I915_READ(TRANS_DDI_FUNC_CTL(TRANSCODER_EDP)); - if (tmp & TRANS_DDI_FUNC_ENABLE) - is_edp = true; - - if (IS_ICELAKE(dev_priv)) { - tmp = I915_READ(TRANS_DDI_FUNC_CTL(TRANSCODER_DSI_0)); - if (tmp & TRANS_DDI_FUNC_ENABLE) - is_dsi = true; - } - - WARN_ON(is_edp && is_dsi); - - if (is_edp || is_dsi) { + for_each_set_bit(panel_transcoder, &panel_transcoder_mask, 32) { enum pipe trans_pipe; + + tmp = I915_READ(TRANS_DDI_FUNC_CTL(panel_transcoder)); + if (!(tmp & TRANS_DDI_FUNC_ENABLE)) + continue; + + /* + * Log all enabled ones, only use the first one. + * + * FIXME: This won't work for two separate DSI displays. + */ + enabled_panel_transcoders |= BIT(panel_transcoder); + if (enabled_panel_transcoders != BIT(panel_transcoder)) + continue; + switch (tmp & TRANS_DDI_EDP_INPUT_MASK) { default: - WARN(1, "unknown pipe linked to edp transcoder\n"); + WARN(1, "unknown pipe linked to transcoder %s\n", + transcoder_name(panel_transcoder)); /* fall through */ case TRANS_DDI_EDP_INPUT_A_ONOFF: case TRANS_DDI_EDP_INPUT_A_ON: @@ -9520,14 +9528,16 @@ static bool hsw_get_transcoder_state(struct intel_crtc *crtc, break; } - if (trans_pipe == crtc->pipe) { - if (is_edp) - pipe_config->cpu_transcoder = TRANSCODER_EDP; - else if (is_dsi) - pipe_config->cpu_transcoder = TRANSCODER_DSI_0; - } + if (trans_pipe == crtc->pipe) + pipe_config->cpu_transcoder = panel_transcoder; } + /* + * Valid combos: none, eDP, DSI0, DSI1, DSI0+DSI1 + */ + WARN_ON((enabled_panel_transcoders & BIT(TRANSCODER_EDP)) && + enabled_panel_transcoders != BIT(TRANSCODER_EDP)); + power_domain = POWER_DOMAIN_TRANSCODER(pipe_config->cpu_transcoder); if (!intel_display_power_get_if_enabled(dev_priv, power_domain)) return false; From 4377d4e0d3d511986033ba7b4182d5a80b7f9ea2 Mon Sep 17 00:00:00 2001 From: Jani Nikula Date: Tue, 4 Dec 2018 19:26:17 +0200 Subject: [PATCH 98/98] drm/i915: Update DRIVER_DATE to 20181204 Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/i915_drv.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index c5f01964f0fb..b1c31967194b 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -90,8 +90,8 @@ #define DRIVER_NAME "i915" #define DRIVER_DESC "Intel Graphics" -#define DRIVER_DATE "20181122" -#define DRIVER_TIMESTAMP 1542898187 +#define DRIVER_DATE "20181204" +#define DRIVER_TIMESTAMP 1543944377 /* Use I915_STATE_WARN(x) and I915_STATE_WARN_ON() (rather than WARN() and * WARN_ON()) for hw state sanity checks to check for unexpected conditions