linux/drivers/gpu/drm/i915/display/intel_modeset_setup.c

1005 lines
29 KiB
C
Raw Permalink Normal View History

// SPDX-License-Identifier: MIT
/*
* Copyright © 2022 Intel Corporation
*
* Read out the current hardware modeset state, and sanitize it to the current
* state.
*/
#include <drm/drm_atomic_state_helper.h>
#include <drm/drm_atomic_uapi.h>
#include <drm/drm_print.h>
#include <drm/drm_vblank.h>
#include "i915_reg.h"
#include "i9xx_wm.h"
#include "intel_atomic.h"
#include "intel_bw.h"
drm/i915/cmtg: Disable the CMTG The CMTG is a timing generator that runs in parallel with transcoders timing generators and can be used as a reference for synchronization. We have observed that we are inheriting from GOP a display configuration with the CMTG enabled. Because our driver doesn't currently implement any CMTG sequences, the CMTG ends up still enabled after our driver takes over. We need to make sure that the CMTG is not enabled if we are not going to use it. For that, let's add a partial implementation in our driver that only cares about disabling the CMTG if it was found enabled during initial hardware readout. In the future, we can also implement sequences for using the CMTG if that becomes a needed feature. For now, we only deal with cases when it is possible to disable the CMTG without requiring a modeset. For earlier display versions, we simply skip if we find the CMTG enabled and we can't disable it without a proper modeset. In the future, we need to properly handle that case. v2: - DG2 does not have the CMTG. Update HAS_CMTG() accordingly. - Update logic to force disabling of CMTG only for initial commit. v3: - Add missing changes for v2 that were staged but not committed. v4: - Avoid if/else duplication in intel_cmtg_dump_state() by using "n/a" for CMTG B enabled/disabled string for platforms without it. (Jani) - Prefer intel_cmtg_readout_hw_state() over intel_cmtg_readout_state(). (Jani) - Use display struct instead of i915 as first parameter for TRANS_DDI_FUNC_CTL2(). (Jani) - Fewer continuation lines in variable declaration/initialization for better readability. (Jani) - Coding style improvements. (Jani) - Use drm_dbg_kms() instead of drm_info() for logging the disabling of the CMTG. - Make struct intel_cmtg_state entirely private to intel_cmtg.c. v5: - Do the disable sequence as part of the sanitization step after hardware readout instead of initial modeset commit. (Jani) - Adapt to commit 15133582465f ("drm/i915/display: convert global state to struct intel_display") by using a display struct instead of i915 as argument for intel_atomic_global_obj_init(). v6: - Do not track CMTG state as a global state. (Ville) - Simplify the driver logic by only disabling the CMTG only on cases when a modeset is not required. (Ville) v7: - Remove the call to drm_WARN_ON() when checking intel_cmtg_disable_requires_modeset() and use a FIXME in the comment instead. - Remove the !HAS_CMTG() guard from intel_cmtg_get_config(), which is static and its caller is already protected by that same condition. - Also take the opportunity to put some Bspec references in the commit trailers section. v8: - Use HAS_TRANSCODER() instead of intel_crtc_for_pipe(). (Ville) - Ensure transcoder power well is enabled before reading TRANS_DDI_FUNC_CTL2. (Ville) Bspec: 68915, 49262 Cc: Jani Nikula <jani.nikula@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20250124173956.46534-1-gustavo.sousa@intel.com
2025-01-24 14:38:14 -03:00
#include "intel_cmtg.h"
#include "intel_color.h"
#include "intel_crtc.h"
#include "intel_crtc_state_dump.h"
#include "intel_ddi.h"
#include "intel_de.h"
#include "intel_display.h"
#include "intel_display_power.h"
#include "intel_display_regs.h"
#include "intel_display_types.h"
#include "intel_dmc.h"
#include "intel_fifo_underrun.h"
#include "intel_modeset_setup.h"
#include "intel_pch_display.h"
drm/i915/mtl: Add support for PM DEMAND MTL introduces a new way to instruct the PUnit with power and bandwidth requirements of DE. Add the functionality to program the registers and handle waits using interrupts. The current wait time for timeouts is programmed for 10 msecs to factor in the worst case scenarios. Changes made to use REG_BIT for a register that we touched(GEN8_DE_MISC_IER _MMIO). Wa_14016740474 is added which applies to Xe_LPD+ display v2: checkpatch warning fixes, simplify program pmdemand part v3: update to dbufs and pipes values to pmdemand register(stan) Removed the macro usage in update_pmdemand_values() v4: move the pmdemand_pre_plane_update before cdclk update pmdemand_needs_update included cdclk params comparisons pmdemand_state NULL check (Gustavo) pmdemand.o in sorted order in the makefile (Jani) update pmdemand misc irq handler loop (Gustavo) active phys bitmask and programming correction (Gustavo) v5: simplify pmdemand_state structure simplify methods to find active phys and max port clock Timeout in case of previou pmdemand task pending (Gustavo) v6: rebasing updates to max_ddiclk calculations (Gustavo) updates to active_phys count method (Gustavo) v7: use two separate loop to iterate throug old and new crtc states to calculate the active phys (Gustavo) v8: use uniform function names (Gustavo) v9: For phys change iterate through connectors (Imre) Look for change in phys for pmdemand update (Gustavo, Imre) Some more stlying changes (Imre) Update pmdemand state during HW readout/sanitize (Imre) v10: Fix CI checkpatch warnings v11: use correct pmdemand object pointer during hw readout, simplify the check for phys need update (Gustavo) v12: Handle possible non serialize cases (Imre) Initialise also pmdemand params HW readout (Imre) Update active phys mask during sanitize calls (Imre) Check TC/encoder changes to limit connector update (Imre) v13: Check display version before accessing pmdemand functions v14: Move is_serialized to intel_global_state.c simplify update params and other stlying issues (Imre) Bspec: 66451, 64636, 64602, 64603 Cc: Matt Atwood <matthew.s.atwood@intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: José Roberto de Souza <jose.souza@intel.com> Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: Mika Kahola <mika.kahola@intel.com> Signed-off-by: Vinod Govindapillai <vinod.govindapillai@intel.com> Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> #v4 Acked-by: Gustavo Sousa <gustavo.sousa@intel.com> #v11 Reviewed-by: Imre Deak <imre.deak@intel.com> [RK: Fixed minor typo in one of the comments. s/qclck_gc/qclk_gv/] Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230606201032.347449-1-vinod.govindapillai@intel.com
2023-06-06 23:10:32 +03:00
#include "intel_pmdemand.h"
drm/i915/tc: Reset TypeC PHYs left enabled in DP-alt mode after the sink disconnects If the output on a DP-alt link with its sink disconnected is kept enabled for too long (about 20 sec), then some IOM/TCSS firmware timeout will cause havoc on the PCI bus, at least for other GFX devices on it which will stop powering up. Since user space is not guaranteed to do a disabling modeset in time, switch such disconnected but active links to TBT mode - which is without such shortcomings - with a 2 second delay. If the above condition is detected already during the driver load/system resume sanitization step disable the output instead, as at that point no user space or kernel client depends on a consistent output state yet and because subsequent atomic modeset on such connectors - without the actual sink capabilities available - can fail. An active/disconnected port as above will also block the HPD status of other active/disconnected ports to get updated (stuck in the connected state), until the former port is disabled, its PHY is disconnected and a ~10 ms delay has elapsed. This means the link state for all TypeC ports/CRTCs must be rechecked after a CRTC is disabled due to the above reason. For this disconnect the PHY synchronously after the CRTC/port is disabled and recheck all CRTCs for the above condition whenever such a port is disabled. To account for a race condition during driver loading where the sink is disconnected after the above sanitization step and before the HPD interrupts get enabled, do an explicit check/link reset if needed from the encoder's late_register hook, which is called after the HPD interrupts are enabled already. v2: - Handle an active/disconnected port blocking the HPD state update of another active/disconnected port. - Cancel the delayed work resetting the link also from the encoder enable/suspend/shutdown hooks. - Rebase on the earlier intel_modeset_lock_ctx_retry() addition, fixing here the missed atomic state reset in case of a retry. - Fix handling of an error return from intel_atomic_get_crtc_state(). - Recheck if the port needs to be reset after all the atomic state is locked and async commits are waited on. v3: - Add intel_crtc_needs_link_reset(), instead of open-coding it, keep intel_crtc_has_encoders(). (Ville) - Fix state dumping and use a bitmask to track disabled CRTCs in intel_sanitize_all_crtcs(). (Ville) - Set internal in intel_atomic_state right after allocating it. (Ville) - Recheck all CRTCs (not yet force-disabled) after a CRTC is force-disabled for any reason (not only due to a link state) in intel_sanitize_all_crtcs(). - Reduce delay after CRTC disabling to 20ms, and use the simpler msleep(). - Clarify code comment about HPD behaviour in intel_sanitize_all_crtcs(). - Move all the TC link reset logic to intel_tc.c . - Cancel the link reset work synchronously during system suspend, driver unload and shutdown. v4: - Rebased on previous patch, which allows calling the TC port suspend/cleanup handlers without modeset locks held; remove the display driver suspended assert from the link reset work accordingly. v5: (Ville) - Remove reset work canceling from intel_ddi_pre_pll_enable(). - Track a crtc vs. pipe mask in intel_sanitize_all_crtcs(). - Add reset_link_commit() to clarify the intel_modeset_lock_ctx_retry loop. Cc: Kai-Heng Feng <kai.heng.feng@canonical.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Tested-by: Kai-Heng Feng <kai.heng.feng@canonical.com> Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/5860 Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230512195513.2699-2-imre.deak@intel.com
2023-05-12 22:55:13 +03:00
#include "intel_tc.h"
#include "intel_vblank.h"
#include "intel_vga.h"
#include "intel_wm.h"
#include "skl_watermark.h"
static void intel_crtc_disable_noatomic_begin(struct intel_crtc *crtc,
struct drm_modeset_acquire_ctx *ctx)
{
struct intel_display *display = to_intel_display(crtc);
struct intel_crtc_state *crtc_state =
to_intel_crtc_state(crtc->base.state);
struct intel_plane *plane;
struct drm_atomic_state *state;
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
struct intel_crtc *temp_crtc;
enum pipe pipe = crtc->pipe;
if (!crtc_state->hw.active)
return;
for_each_intel_plane_on_crtc(display->drm, crtc, plane) {
const struct intel_plane_state *plane_state =
to_intel_plane_state(plane->base.state);
if (plane_state->uapi.visible)
intel_plane_disable_noatomic(crtc, plane);
}
state = drm_atomic_state_alloc(display->drm);
if (!state) {
drm_dbg_kms(display->drm,
"failed to disable [CRTC:%d:%s], out of memory",
crtc->base.base.id, crtc->base.name);
return;
}
state->acquire_ctx = ctx;
to_intel_atomic_state(state)->internal = true;
/* Everything's already locked, -EDEADLK can't happen. */
for_each_intel_crtc_in_pipe_mask(display->drm, temp_crtc,
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
BIT(pipe) |
intel_crtc_joiner_secondary_pipes(crtc_state)) {
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
struct intel_crtc_state *temp_crtc_state =
intel_atomic_get_crtc_state(state, temp_crtc);
int ret;
ret = drm_atomic_add_affected_connectors(state, &temp_crtc->base);
drm_WARN_ON(display->drm, IS_ERR(temp_crtc_state) || ret);
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
}
display->funcs.display->crtc_disable(to_intel_atomic_state(state), crtc);
drm_atomic_state_put(state);
drm_dbg_kms(display->drm,
"[CRTC:%d:%s] hw state adjusted, was enabled, now disabled\n",
crtc->base.base.id, crtc->base.name);
crtc->active = false;
crtc->base.enabled = false;
if (crtc_state->intel_dpll)
intel_dpll_crtc_put(crtc,
crtc_state->intel_dpll,
&crtc_state->intel_dpll->state);
}
static void set_encoder_for_connector(struct intel_connector *connector,
struct intel_encoder *encoder)
{
struct drm_connector_state *conn_state = connector->base.state;
if (conn_state->crtc)
drm_connector_put(&connector->base);
if (encoder) {
conn_state->best_encoder = &encoder->base;
conn_state->crtc = encoder->base.crtc;
drm_connector_get(&connector->base);
} else {
conn_state->best_encoder = NULL;
conn_state->crtc = NULL;
}
}
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
static void reset_encoder_connector_state(struct intel_encoder *encoder)
{
struct intel_display *display = to_intel_display(encoder);
drm/i915/mtl: Add support for PM DEMAND MTL introduces a new way to instruct the PUnit with power and bandwidth requirements of DE. Add the functionality to program the registers and handle waits using interrupts. The current wait time for timeouts is programmed for 10 msecs to factor in the worst case scenarios. Changes made to use REG_BIT for a register that we touched(GEN8_DE_MISC_IER _MMIO). Wa_14016740474 is added which applies to Xe_LPD+ display v2: checkpatch warning fixes, simplify program pmdemand part v3: update to dbufs and pipes values to pmdemand register(stan) Removed the macro usage in update_pmdemand_values() v4: move the pmdemand_pre_plane_update before cdclk update pmdemand_needs_update included cdclk params comparisons pmdemand_state NULL check (Gustavo) pmdemand.o in sorted order in the makefile (Jani) update pmdemand misc irq handler loop (Gustavo) active phys bitmask and programming correction (Gustavo) v5: simplify pmdemand_state structure simplify methods to find active phys and max port clock Timeout in case of previou pmdemand task pending (Gustavo) v6: rebasing updates to max_ddiclk calculations (Gustavo) updates to active_phys count method (Gustavo) v7: use two separate loop to iterate throug old and new crtc states to calculate the active phys (Gustavo) v8: use uniform function names (Gustavo) v9: For phys change iterate through connectors (Imre) Look for change in phys for pmdemand update (Gustavo, Imre) Some more stlying changes (Imre) Update pmdemand state during HW readout/sanitize (Imre) v10: Fix CI checkpatch warnings v11: use correct pmdemand object pointer during hw readout, simplify the check for phys need update (Gustavo) v12: Handle possible non serialize cases (Imre) Initialise also pmdemand params HW readout (Imre) Update active phys mask during sanitize calls (Imre) Check TC/encoder changes to limit connector update (Imre) v13: Check display version before accessing pmdemand functions v14: Move is_serialized to intel_global_state.c simplify update params and other stlying issues (Imre) Bspec: 66451, 64636, 64602, 64603 Cc: Matt Atwood <matthew.s.atwood@intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: José Roberto de Souza <jose.souza@intel.com> Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: Mika Kahola <mika.kahola@intel.com> Signed-off-by: Vinod Govindapillai <vinod.govindapillai@intel.com> Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> #v4 Acked-by: Gustavo Sousa <gustavo.sousa@intel.com> #v11 Reviewed-by: Imre Deak <imre.deak@intel.com> [RK: Fixed minor typo in one of the comments. s/qclck_gc/qclk_gv/] Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230606201032.347449-1-vinod.govindapillai@intel.com
2023-06-06 23:10:32 +03:00
struct intel_pmdemand_state *pmdemand_state =
to_intel_pmdemand_state(display->pmdemand.obj.state);
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
struct intel_connector *connector;
struct drm_connector_list_iter conn_iter;
drm_connector_list_iter_begin(display->drm, &conn_iter);
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
for_each_intel_connector_iter(connector, &conn_iter) {
if (connector->base.encoder != &encoder->base)
continue;
drm/i915/mtl: Add support for PM DEMAND MTL introduces a new way to instruct the PUnit with power and bandwidth requirements of DE. Add the functionality to program the registers and handle waits using interrupts. The current wait time for timeouts is programmed for 10 msecs to factor in the worst case scenarios. Changes made to use REG_BIT for a register that we touched(GEN8_DE_MISC_IER _MMIO). Wa_14016740474 is added which applies to Xe_LPD+ display v2: checkpatch warning fixes, simplify program pmdemand part v3: update to dbufs and pipes values to pmdemand register(stan) Removed the macro usage in update_pmdemand_values() v4: move the pmdemand_pre_plane_update before cdclk update pmdemand_needs_update included cdclk params comparisons pmdemand_state NULL check (Gustavo) pmdemand.o in sorted order in the makefile (Jani) update pmdemand misc irq handler loop (Gustavo) active phys bitmask and programming correction (Gustavo) v5: simplify pmdemand_state structure simplify methods to find active phys and max port clock Timeout in case of previou pmdemand task pending (Gustavo) v6: rebasing updates to max_ddiclk calculations (Gustavo) updates to active_phys count method (Gustavo) v7: use two separate loop to iterate throug old and new crtc states to calculate the active phys (Gustavo) v8: use uniform function names (Gustavo) v9: For phys change iterate through connectors (Imre) Look for change in phys for pmdemand update (Gustavo, Imre) Some more stlying changes (Imre) Update pmdemand state during HW readout/sanitize (Imre) v10: Fix CI checkpatch warnings v11: use correct pmdemand object pointer during hw readout, simplify the check for phys need update (Gustavo) v12: Handle possible non serialize cases (Imre) Initialise also pmdemand params HW readout (Imre) Update active phys mask during sanitize calls (Imre) Check TC/encoder changes to limit connector update (Imre) v13: Check display version before accessing pmdemand functions v14: Move is_serialized to intel_global_state.c simplify update params and other stlying issues (Imre) Bspec: 66451, 64636, 64602, 64603 Cc: Matt Atwood <matthew.s.atwood@intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: José Roberto de Souza <jose.souza@intel.com> Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: Mika Kahola <mika.kahola@intel.com> Signed-off-by: Vinod Govindapillai <vinod.govindapillai@intel.com> Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> #v4 Acked-by: Gustavo Sousa <gustavo.sousa@intel.com> #v11 Reviewed-by: Imre Deak <imre.deak@intel.com> [RK: Fixed minor typo in one of the comments. s/qclck_gc/qclk_gv/] Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230606201032.347449-1-vinod.govindapillai@intel.com
2023-06-06 23:10:32 +03:00
/* Clear the corresponding bit in pmdemand active phys mask */
intel_pmdemand_update_phys_mask(display, encoder,
drm/i915/mtl: Add support for PM DEMAND MTL introduces a new way to instruct the PUnit with power and bandwidth requirements of DE. Add the functionality to program the registers and handle waits using interrupts. The current wait time for timeouts is programmed for 10 msecs to factor in the worst case scenarios. Changes made to use REG_BIT for a register that we touched(GEN8_DE_MISC_IER _MMIO). Wa_14016740474 is added which applies to Xe_LPD+ display v2: checkpatch warning fixes, simplify program pmdemand part v3: update to dbufs and pipes values to pmdemand register(stan) Removed the macro usage in update_pmdemand_values() v4: move the pmdemand_pre_plane_update before cdclk update pmdemand_needs_update included cdclk params comparisons pmdemand_state NULL check (Gustavo) pmdemand.o in sorted order in the makefile (Jani) update pmdemand misc irq handler loop (Gustavo) active phys bitmask and programming correction (Gustavo) v5: simplify pmdemand_state structure simplify methods to find active phys and max port clock Timeout in case of previou pmdemand task pending (Gustavo) v6: rebasing updates to max_ddiclk calculations (Gustavo) updates to active_phys count method (Gustavo) v7: use two separate loop to iterate throug old and new crtc states to calculate the active phys (Gustavo) v8: use uniform function names (Gustavo) v9: For phys change iterate through connectors (Imre) Look for change in phys for pmdemand update (Gustavo, Imre) Some more stlying changes (Imre) Update pmdemand state during HW readout/sanitize (Imre) v10: Fix CI checkpatch warnings v11: use correct pmdemand object pointer during hw readout, simplify the check for phys need update (Gustavo) v12: Handle possible non serialize cases (Imre) Initialise also pmdemand params HW readout (Imre) Update active phys mask during sanitize calls (Imre) Check TC/encoder changes to limit connector update (Imre) v13: Check display version before accessing pmdemand functions v14: Move is_serialized to intel_global_state.c simplify update params and other stlying issues (Imre) Bspec: 66451, 64636, 64602, 64603 Cc: Matt Atwood <matthew.s.atwood@intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: José Roberto de Souza <jose.souza@intel.com> Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: Mika Kahola <mika.kahola@intel.com> Signed-off-by: Vinod Govindapillai <vinod.govindapillai@intel.com> Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> #v4 Acked-by: Gustavo Sousa <gustavo.sousa@intel.com> #v11 Reviewed-by: Imre Deak <imre.deak@intel.com> [RK: Fixed minor typo in one of the comments. s/qclck_gc/qclk_gv/] Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230606201032.347449-1-vinod.govindapillai@intel.com
2023-06-06 23:10:32 +03:00
pmdemand_state, false);
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
set_encoder_for_connector(connector, NULL);
connector->base.dpms = DRM_MODE_DPMS_OFF;
connector->base.encoder = NULL;
}
drm_connector_list_iter_end(&conn_iter);
}
static void reset_crtc_encoder_state(struct intel_crtc *crtc)
{
struct intel_display *display = to_intel_display(crtc);
struct intel_encoder *encoder;
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
for_each_encoder_on_crtc(display->drm, &crtc->base, encoder) {
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
reset_encoder_connector_state(encoder);
encoder->base.crtc = NULL;
}
}
static void intel_crtc_disable_noatomic_complete(struct intel_crtc *crtc)
{
struct intel_display *display = to_intel_display(crtc);
drm/i915/mtl: Add support for PM DEMAND MTL introduces a new way to instruct the PUnit with power and bandwidth requirements of DE. Add the functionality to program the registers and handle waits using interrupts. The current wait time for timeouts is programmed for 10 msecs to factor in the worst case scenarios. Changes made to use REG_BIT for a register that we touched(GEN8_DE_MISC_IER _MMIO). Wa_14016740474 is added which applies to Xe_LPD+ display v2: checkpatch warning fixes, simplify program pmdemand part v3: update to dbufs and pipes values to pmdemand register(stan) Removed the macro usage in update_pmdemand_values() v4: move the pmdemand_pre_plane_update before cdclk update pmdemand_needs_update included cdclk params comparisons pmdemand_state NULL check (Gustavo) pmdemand.o in sorted order in the makefile (Jani) update pmdemand misc irq handler loop (Gustavo) active phys bitmask and programming correction (Gustavo) v5: simplify pmdemand_state structure simplify methods to find active phys and max port clock Timeout in case of previou pmdemand task pending (Gustavo) v6: rebasing updates to max_ddiclk calculations (Gustavo) updates to active_phys count method (Gustavo) v7: use two separate loop to iterate throug old and new crtc states to calculate the active phys (Gustavo) v8: use uniform function names (Gustavo) v9: For phys change iterate through connectors (Imre) Look for change in phys for pmdemand update (Gustavo, Imre) Some more stlying changes (Imre) Update pmdemand state during HW readout/sanitize (Imre) v10: Fix CI checkpatch warnings v11: use correct pmdemand object pointer during hw readout, simplify the check for phys need update (Gustavo) v12: Handle possible non serialize cases (Imre) Initialise also pmdemand params HW readout (Imre) Update active phys mask during sanitize calls (Imre) Check TC/encoder changes to limit connector update (Imre) v13: Check display version before accessing pmdemand functions v14: Move is_serialized to intel_global_state.c simplify update params and other stlying issues (Imre) Bspec: 66451, 64636, 64602, 64603 Cc: Matt Atwood <matthew.s.atwood@intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: José Roberto de Souza <jose.souza@intel.com> Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: Mika Kahola <mika.kahola@intel.com> Signed-off-by: Vinod Govindapillai <vinod.govindapillai@intel.com> Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> #v4 Acked-by: Gustavo Sousa <gustavo.sousa@intel.com> #v11 Reviewed-by: Imre Deak <imre.deak@intel.com> [RK: Fixed minor typo in one of the comments. s/qclck_gc/qclk_gv/] Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230606201032.347449-1-vinod.govindapillai@intel.com
2023-06-06 23:10:32 +03:00
struct intel_pmdemand_state *pmdemand_state =
to_intel_pmdemand_state(display->pmdemand.obj.state);
struct intel_crtc_state *crtc_state =
to_intel_crtc_state(crtc->base.state);
enum pipe pipe = crtc->pipe;
__drm_atomic_helper_crtc_destroy_state(&crtc_state->uapi);
intel_crtc_free_hw_state(crtc_state);
intel_crtc_state_reset(crtc_state, crtc);
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
reset_crtc_encoder_state(crtc);
intel_fbc_disable(crtc);
intel_update_watermarks(display);
intel_display_power_put_all_in_set(display, &crtc->enabled_power_domains);
intel_cdclk_crtc_disable_noatomic(crtc);
skl_wm_crtc_disable_noatomic(crtc);
intel_bw_crtc_disable_noatomic(crtc);
drm/i915/mtl: Add support for PM DEMAND MTL introduces a new way to instruct the PUnit with power and bandwidth requirements of DE. Add the functionality to program the registers and handle waits using interrupts. The current wait time for timeouts is programmed for 10 msecs to factor in the worst case scenarios. Changes made to use REG_BIT for a register that we touched(GEN8_DE_MISC_IER _MMIO). Wa_14016740474 is added which applies to Xe_LPD+ display v2: checkpatch warning fixes, simplify program pmdemand part v3: update to dbufs and pipes values to pmdemand register(stan) Removed the macro usage in update_pmdemand_values() v4: move the pmdemand_pre_plane_update before cdclk update pmdemand_needs_update included cdclk params comparisons pmdemand_state NULL check (Gustavo) pmdemand.o in sorted order in the makefile (Jani) update pmdemand misc irq handler loop (Gustavo) active phys bitmask and programming correction (Gustavo) v5: simplify pmdemand_state structure simplify methods to find active phys and max port clock Timeout in case of previou pmdemand task pending (Gustavo) v6: rebasing updates to max_ddiclk calculations (Gustavo) updates to active_phys count method (Gustavo) v7: use two separate loop to iterate throug old and new crtc states to calculate the active phys (Gustavo) v8: use uniform function names (Gustavo) v9: For phys change iterate through connectors (Imre) Look for change in phys for pmdemand update (Gustavo, Imre) Some more stlying changes (Imre) Update pmdemand state during HW readout/sanitize (Imre) v10: Fix CI checkpatch warnings v11: use correct pmdemand object pointer during hw readout, simplify the check for phys need update (Gustavo) v12: Handle possible non serialize cases (Imre) Initialise also pmdemand params HW readout (Imre) Update active phys mask during sanitize calls (Imre) Check TC/encoder changes to limit connector update (Imre) v13: Check display version before accessing pmdemand functions v14: Move is_serialized to intel_global_state.c simplify update params and other stlying issues (Imre) Bspec: 66451, 64636, 64602, 64603 Cc: Matt Atwood <matthew.s.atwood@intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: José Roberto de Souza <jose.souza@intel.com> Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: Mika Kahola <mika.kahola@intel.com> Signed-off-by: Vinod Govindapillai <vinod.govindapillai@intel.com> Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> #v4 Acked-by: Gustavo Sousa <gustavo.sousa@intel.com> #v11 Reviewed-by: Imre Deak <imre.deak@intel.com> [RK: Fixed minor typo in one of the comments. s/qclck_gc/qclk_gv/] Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230606201032.347449-1-vinod.govindapillai@intel.com
2023-06-06 23:10:32 +03:00
intel_pmdemand_update_port_clock(display, pmdemand_state, pipe, 0);
}
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
/*
* Return all the pipes using a transcoder in @transcoder_mask.
* For joiner configs return only the joiner primary.
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
*/
static u8 get_transcoder_pipes(struct intel_display *display,
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
u8 transcoder_mask)
{
struct intel_crtc *temp_crtc;
u8 pipes = 0;
for_each_intel_crtc(display->drm, temp_crtc) {
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
struct intel_crtc_state *temp_crtc_state =
to_intel_crtc_state(temp_crtc->base.state);
if (temp_crtc_state->cpu_transcoder == INVALID_TRANSCODER)
continue;
if (intel_crtc_is_joiner_secondary(temp_crtc_state))
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
continue;
if (transcoder_mask & BIT(temp_crtc_state->cpu_transcoder))
pipes |= BIT(temp_crtc->pipe);
}
return pipes;
}
/*
* Return the port sync master and slave pipes linked to @crtc.
* For joiner configs return only the joiner primary pipes.
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
*/
static void get_portsync_pipes(struct intel_crtc *crtc,
u8 *master_pipe_mask, u8 *slave_pipes_mask)
{
struct intel_display *display = to_intel_display(crtc);
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
struct intel_crtc_state *crtc_state =
to_intel_crtc_state(crtc->base.state);
struct intel_crtc *master_crtc;
struct intel_crtc_state *master_crtc_state;
enum transcoder master_transcoder;
if (!is_trans_port_sync_mode(crtc_state)) {
*master_pipe_mask = BIT(crtc->pipe);
*slave_pipes_mask = 0;
return;
}
if (is_trans_port_sync_master(crtc_state))
master_transcoder = crtc_state->cpu_transcoder;
else
master_transcoder = crtc_state->master_transcoder;
*master_pipe_mask = get_transcoder_pipes(display, BIT(master_transcoder));
drm_WARN_ON(display->drm, !is_power_of_2(*master_pipe_mask));
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
master_crtc = intel_crtc_for_pipe(display, ffs(*master_pipe_mask) - 1);
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
master_crtc_state = to_intel_crtc_state(master_crtc->base.state);
*slave_pipes_mask = get_transcoder_pipes(display, master_crtc_state->sync_mode_slaves_mask);
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
}
static u8 get_joiner_secondary_pipes(struct intel_display *display, u8 primary_pipes_mask)
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
{
struct intel_crtc *primary_crtc;
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
u8 pipes = 0;
for_each_intel_crtc_in_pipe_mask(display->drm, primary_crtc, primary_pipes_mask) {
struct intel_crtc_state *primary_crtc_state =
to_intel_crtc_state(primary_crtc->base.state);
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
pipes |= intel_crtc_joiner_secondary_pipes(primary_crtc_state);
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
}
return pipes;
}
static void intel_crtc_disable_noatomic(struct intel_crtc *crtc,
struct drm_modeset_acquire_ctx *ctx)
{
struct intel_display *display = to_intel_display(crtc);
struct intel_crtc *temp_crtc;
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
u8 portsync_master_mask;
u8 portsync_slaves_mask;
u8 joiner_secondaries_mask;
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
/* TODO: Add support for MST */
get_portsync_pipes(crtc, &portsync_master_mask, &portsync_slaves_mask);
joiner_secondaries_mask = get_joiner_secondary_pipes(display,
portsync_master_mask |
portsync_slaves_mask);
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
drm_WARN_ON(display->drm,
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
portsync_master_mask & portsync_slaves_mask ||
portsync_master_mask & joiner_secondaries_mask ||
portsync_slaves_mask & joiner_secondaries_mask);
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
for_each_intel_crtc_in_pipe_mask(display->drm, temp_crtc, joiner_secondaries_mask)
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
intel_crtc_disable_noatomic_begin(temp_crtc, ctx);
for_each_intel_crtc_in_pipe_mask(display->drm, temp_crtc, portsync_slaves_mask)
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
intel_crtc_disable_noatomic_begin(temp_crtc, ctx);
for_each_intel_crtc_in_pipe_mask(display->drm, temp_crtc, portsync_master_mask)
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
intel_crtc_disable_noatomic_begin(temp_crtc, ctx);
for_each_intel_crtc_in_pipe_mask(display->drm, temp_crtc,
joiner_secondaries_mask |
drm/i915: Add support for disabling any CRTCs during HW readout/sanitization During HW readout/sanitization CRTCs can be disabled only if they don't have an attached encoder (and so the encoder disable hooks don't need to be called). An upcoming patch will need to disable CRTCs also with an attached encoder, so add support for this. For bigjoiner configs the encoder disabling hooks require the slave CRTC states, so add these too to the atomic state. Since the connector atomic state is already up-to-date when the CRTC is disabled the connector state needs to be updated (reset) after the CRTC is disabled, make this so. Follow the proper order of disabling first all bigjoiner slaves, then any port synced CRTC slaves followed by the CRTC originally requested to be disabled. v2: - Fix calculating the bigjoiner_masters mask in a port sync config, (Ville) - Keep _noatomic suffix in intel_crtc_disable_noatomic(). (Ville) - Rebase on full CRTC state reset in this patchset, not requiring resetting the bigjoiner state separately and (instead) resetting the full atomic CRTC and related global state after all linked pipes got disabled. - Disable portsync slaves before a portsync master. - Disable a portsync master if a linked portsync slave is disabled. v3: (Ville) - Use s/u32/u8 for transcoder and pipe masks. - Use is_power_of_2() instead of hweight()==1. Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230510103131.1618266-8-imre.deak@intel.com
2023-05-10 13:31:24 +03:00
portsync_slaves_mask |
portsync_master_mask)
intel_crtc_disable_noatomic_complete(temp_crtc);
}
static void intel_modeset_update_connector_atomic_state(struct intel_display *display)
{
struct intel_connector *connector;
struct drm_connector_list_iter conn_iter;
drm_connector_list_iter_begin(display->drm, &conn_iter);
for_each_intel_connector_iter(connector, &conn_iter) {
struct drm_connector_state *conn_state = connector->base.state;
struct intel_encoder *encoder =
to_intel_encoder(connector->base.encoder);
set_encoder_for_connector(connector, encoder);
if (encoder) {
struct intel_crtc *crtc =
to_intel_crtc(encoder->base.crtc);
const struct intel_crtc_state *crtc_state =
to_intel_crtc_state(crtc->base.state);
conn_state->max_bpc = (crtc_state->pipe_bpp ?: 24) / 3;
}
}
drm_connector_list_iter_end(&conn_iter);
}
static void intel_crtc_copy_hw_to_uapi_state(struct intel_crtc_state *crtc_state)
{
struct intel_display *display = to_intel_display(crtc_state);
if (intel_crtc_is_joiner_secondary(crtc_state))
return;
crtc_state->uapi.enable = crtc_state->hw.enable;
crtc_state->uapi.active = crtc_state->hw.active;
drm_WARN_ON(crtc_state->uapi.crtc->dev,
drm_atomic_set_mode_for_crtc(&crtc_state->uapi, &crtc_state->hw.mode) < 0);
crtc_state->uapi.adjusted_mode = crtc_state->hw.adjusted_mode;
crtc_state->uapi.scaling_filter = crtc_state->hw.scaling_filter;
if (DISPLAY_INFO(display)->color.degamma_lut_size) {
/* assume 1:1 mapping */
drm_property_replace_blob(&crtc_state->hw.degamma_lut,
crtc_state->pre_csc_lut);
drm_property_replace_blob(&crtc_state->hw.gamma_lut,
crtc_state->post_csc_lut);
} else {
/*
* ilk/snb hw may be configured for either pre_csc_lut
* or post_csc_lut, but we don't advertise degamma_lut as
* being available in the uapi since there is only one
* hardware LUT. Always assign the result of the readout
* to gamma_lut as that is the only valid source of LUTs
* in the uapi.
*/
drm_WARN_ON(display->drm, crtc_state->post_csc_lut &&
crtc_state->pre_csc_lut);
drm_property_replace_blob(&crtc_state->hw.degamma_lut,
NULL);
drm_property_replace_blob(&crtc_state->hw.gamma_lut,
crtc_state->post_csc_lut ?:
crtc_state->pre_csc_lut);
}
drm/i915: Introduce crtc_state->{pre,post}_csc_lut Add an extra remapping step between the logical state of the LUTs (hw.(de)gamma_lut) as specified via uapi/bigjoiner copy vs. the actual state of the LUTs programmed into the hardware. With this we should be finally able finish the (de)gamma readout/state checker support for the remaining platforms (ilk-skl) where the same hardware LUT can be positioned either before or after the pipe CSC unit. Where we position it depends on factors such as presence of the logical degamma LUT, RGB vs. YCbCr output, full vs. limited RGB quantization range. Without the extra remapping step the state readout doesn't really know whether the LUT read from the hardware is the degamma or gamma LUT, and so we is unable to accurately store it into our crtc state. With the remapping step we know exactly where to put it given the order of the LUT vs. CSC in the hardware state. Only the initial hw->uapi state readout done during driver load/resume still has the problem of not really knowing what to do with the LUT(s). But we can just assume 1:1 mapping there and let subsequent commits fix things up. Another benefit is that we now have a place for purely internal LUTs, without complicating the bigjoiner uapi->hw copy logic. This should prove useful for streamlining glk degamma LUT handling. Reviewed-by: Uma Shankar <uma.shankar@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221024161514.5340-3-ville.syrjala@linux.intel.com
2022-10-24 19:15:11 +03:00
drm_property_replace_blob(&crtc_state->uapi.degamma_lut,
crtc_state->hw.degamma_lut);
drm_property_replace_blob(&crtc_state->uapi.gamma_lut,
crtc_state->hw.gamma_lut);
drm_property_replace_blob(&crtc_state->uapi.ctm,
crtc_state->hw.ctm);
}
static void
intel_sanitize_plane_mapping(struct intel_display *display)
{
struct intel_crtc *crtc;
if (DISPLAY_VER(display) >= 4)
return;
for_each_intel_crtc(display->drm, crtc) {
struct intel_plane *plane =
to_intel_plane(crtc->base.primary);
struct intel_crtc *plane_crtc;
enum pipe pipe;
if (!plane->get_hw_state(plane, &pipe))
continue;
if (pipe == crtc->pipe)
continue;
drm_dbg_kms(display->drm,
"[PLANE:%d:%s] attached to the wrong pipe, disabling plane\n",
plane->base.base.id, plane->base.name);
plane_crtc = intel_crtc_for_pipe(display, pipe);
intel_plane_disable_noatomic(plane_crtc, plane);
}
}
static bool intel_crtc_has_encoders(struct intel_crtc *crtc)
{
struct drm_device *dev = crtc->base.dev;
struct intel_encoder *encoder;
for_each_encoder_on_crtc(dev, &crtc->base, encoder)
return true;
return false;
}
drm/i915/tc: Reset TypeC PHYs left enabled in DP-alt mode after the sink disconnects If the output on a DP-alt link with its sink disconnected is kept enabled for too long (about 20 sec), then some IOM/TCSS firmware timeout will cause havoc on the PCI bus, at least for other GFX devices on it which will stop powering up. Since user space is not guaranteed to do a disabling modeset in time, switch such disconnected but active links to TBT mode - which is without such shortcomings - with a 2 second delay. If the above condition is detected already during the driver load/system resume sanitization step disable the output instead, as at that point no user space or kernel client depends on a consistent output state yet and because subsequent atomic modeset on such connectors - without the actual sink capabilities available - can fail. An active/disconnected port as above will also block the HPD status of other active/disconnected ports to get updated (stuck in the connected state), until the former port is disabled, its PHY is disconnected and a ~10 ms delay has elapsed. This means the link state for all TypeC ports/CRTCs must be rechecked after a CRTC is disabled due to the above reason. For this disconnect the PHY synchronously after the CRTC/port is disabled and recheck all CRTCs for the above condition whenever such a port is disabled. To account for a race condition during driver loading where the sink is disconnected after the above sanitization step and before the HPD interrupts get enabled, do an explicit check/link reset if needed from the encoder's late_register hook, which is called after the HPD interrupts are enabled already. v2: - Handle an active/disconnected port blocking the HPD state update of another active/disconnected port. - Cancel the delayed work resetting the link also from the encoder enable/suspend/shutdown hooks. - Rebase on the earlier intel_modeset_lock_ctx_retry() addition, fixing here the missed atomic state reset in case of a retry. - Fix handling of an error return from intel_atomic_get_crtc_state(). - Recheck if the port needs to be reset after all the atomic state is locked and async commits are waited on. v3: - Add intel_crtc_needs_link_reset(), instead of open-coding it, keep intel_crtc_has_encoders(). (Ville) - Fix state dumping and use a bitmask to track disabled CRTCs in intel_sanitize_all_crtcs(). (Ville) - Set internal in intel_atomic_state right after allocating it. (Ville) - Recheck all CRTCs (not yet force-disabled) after a CRTC is force-disabled for any reason (not only due to a link state) in intel_sanitize_all_crtcs(). - Reduce delay after CRTC disabling to 20ms, and use the simpler msleep(). - Clarify code comment about HPD behaviour in intel_sanitize_all_crtcs(). - Move all the TC link reset logic to intel_tc.c . - Cancel the link reset work synchronously during system suspend, driver unload and shutdown. v4: - Rebased on previous patch, which allows calling the TC port suspend/cleanup handlers without modeset locks held; remove the display driver suspended assert from the link reset work accordingly. v5: (Ville) - Remove reset work canceling from intel_ddi_pre_pll_enable(). - Track a crtc vs. pipe mask in intel_sanitize_all_crtcs(). - Add reset_link_commit() to clarify the intel_modeset_lock_ctx_retry loop. Cc: Kai-Heng Feng <kai.heng.feng@canonical.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Tested-by: Kai-Heng Feng <kai.heng.feng@canonical.com> Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/5860 Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230512195513.2699-2-imre.deak@intel.com
2023-05-12 22:55:13 +03:00
static bool intel_crtc_needs_link_reset(struct intel_crtc *crtc)
{
struct drm_device *dev = crtc->base.dev;
struct intel_encoder *encoder;
for_each_encoder_on_crtc(dev, &crtc->base, encoder) {
struct intel_digital_port *dig_port = enc_to_dig_port(encoder);
if (dig_port && intel_tc_port_link_needs_reset(dig_port))
return true;
}
return false;
}
static struct intel_connector *intel_encoder_find_connector(struct intel_encoder *encoder)
{
struct intel_display *display = to_intel_display(encoder);
struct drm_connector_list_iter conn_iter;
struct intel_connector *connector;
struct intel_connector *found_connector = NULL;
drm_connector_list_iter_begin(display->drm, &conn_iter);
for_each_intel_connector_iter(connector, &conn_iter) {
if (&encoder->base == connector->base.encoder) {
found_connector = connector;
break;
}
}
drm_connector_list_iter_end(&conn_iter);
return found_connector;
}
static void intel_sanitize_fifo_underrun_reporting(const struct intel_crtc_state *crtc_state)
{
struct intel_display *display = to_intel_display(crtc_state);
struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
/*
* We start out with underrun reporting disabled on active
* pipes to avoid races.
*
* Also on gmch platforms we dont have any hardware bits to
* disable the underrun reporting. Which means we need to start
* out with underrun reporting disabled also on inactive pipes,
* since otherwise we'll complain about the garbage we read when
* e.g. coming up after runtime pm.
*
* No protection against concurrent access is required - at
* worst a fifo underrun happens which also sets this to false.
*/
intel_init_fifo_underrun_reporting(display, crtc,
!crtc_state->hw.active &&
!HAS_GMCH(display));
}
drm/i915/tc: Reset TypeC PHYs left enabled in DP-alt mode after the sink disconnects If the output on a DP-alt link with its sink disconnected is kept enabled for too long (about 20 sec), then some IOM/TCSS firmware timeout will cause havoc on the PCI bus, at least for other GFX devices on it which will stop powering up. Since user space is not guaranteed to do a disabling modeset in time, switch such disconnected but active links to TBT mode - which is without such shortcomings - with a 2 second delay. If the above condition is detected already during the driver load/system resume sanitization step disable the output instead, as at that point no user space or kernel client depends on a consistent output state yet and because subsequent atomic modeset on such connectors - without the actual sink capabilities available - can fail. An active/disconnected port as above will also block the HPD status of other active/disconnected ports to get updated (stuck in the connected state), until the former port is disabled, its PHY is disconnected and a ~10 ms delay has elapsed. This means the link state for all TypeC ports/CRTCs must be rechecked after a CRTC is disabled due to the above reason. For this disconnect the PHY synchronously after the CRTC/port is disabled and recheck all CRTCs for the above condition whenever such a port is disabled. To account for a race condition during driver loading where the sink is disconnected after the above sanitization step and before the HPD interrupts get enabled, do an explicit check/link reset if needed from the encoder's late_register hook, which is called after the HPD interrupts are enabled already. v2: - Handle an active/disconnected port blocking the HPD state update of another active/disconnected port. - Cancel the delayed work resetting the link also from the encoder enable/suspend/shutdown hooks. - Rebase on the earlier intel_modeset_lock_ctx_retry() addition, fixing here the missed atomic state reset in case of a retry. - Fix handling of an error return from intel_atomic_get_crtc_state(). - Recheck if the port needs to be reset after all the atomic state is locked and async commits are waited on. v3: - Add intel_crtc_needs_link_reset(), instead of open-coding it, keep intel_crtc_has_encoders(). (Ville) - Fix state dumping and use a bitmask to track disabled CRTCs in intel_sanitize_all_crtcs(). (Ville) - Set internal in intel_atomic_state right after allocating it. (Ville) - Recheck all CRTCs (not yet force-disabled) after a CRTC is force-disabled for any reason (not only due to a link state) in intel_sanitize_all_crtcs(). - Reduce delay after CRTC disabling to 20ms, and use the simpler msleep(). - Clarify code comment about HPD behaviour in intel_sanitize_all_crtcs(). - Move all the TC link reset logic to intel_tc.c . - Cancel the link reset work synchronously during system suspend, driver unload and shutdown. v4: - Rebased on previous patch, which allows calling the TC port suspend/cleanup handlers without modeset locks held; remove the display driver suspended assert from the link reset work accordingly. v5: (Ville) - Remove reset work canceling from intel_ddi_pre_pll_enable(). - Track a crtc vs. pipe mask in intel_sanitize_all_crtcs(). - Add reset_link_commit() to clarify the intel_modeset_lock_ctx_retry loop. Cc: Kai-Heng Feng <kai.heng.feng@canonical.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Tested-by: Kai-Heng Feng <kai.heng.feng@canonical.com> Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/5860 Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230512195513.2699-2-imre.deak@intel.com
2023-05-12 22:55:13 +03:00
static bool intel_sanitize_crtc(struct intel_crtc *crtc,
struct drm_modeset_acquire_ctx *ctx)
{
struct intel_display *display = to_intel_display(crtc);
struct intel_crtc_state *crtc_state = to_intel_crtc_state(crtc->base.state);
drm/i915/tc: Reset TypeC PHYs left enabled in DP-alt mode after the sink disconnects If the output on a DP-alt link with its sink disconnected is kept enabled for too long (about 20 sec), then some IOM/TCSS firmware timeout will cause havoc on the PCI bus, at least for other GFX devices on it which will stop powering up. Since user space is not guaranteed to do a disabling modeset in time, switch such disconnected but active links to TBT mode - which is without such shortcomings - with a 2 second delay. If the above condition is detected already during the driver load/system resume sanitization step disable the output instead, as at that point no user space or kernel client depends on a consistent output state yet and because subsequent atomic modeset on such connectors - without the actual sink capabilities available - can fail. An active/disconnected port as above will also block the HPD status of other active/disconnected ports to get updated (stuck in the connected state), until the former port is disabled, its PHY is disconnected and a ~10 ms delay has elapsed. This means the link state for all TypeC ports/CRTCs must be rechecked after a CRTC is disabled due to the above reason. For this disconnect the PHY synchronously after the CRTC/port is disabled and recheck all CRTCs for the above condition whenever such a port is disabled. To account for a race condition during driver loading where the sink is disconnected after the above sanitization step and before the HPD interrupts get enabled, do an explicit check/link reset if needed from the encoder's late_register hook, which is called after the HPD interrupts are enabled already. v2: - Handle an active/disconnected port blocking the HPD state update of another active/disconnected port. - Cancel the delayed work resetting the link also from the encoder enable/suspend/shutdown hooks. - Rebase on the earlier intel_modeset_lock_ctx_retry() addition, fixing here the missed atomic state reset in case of a retry. - Fix handling of an error return from intel_atomic_get_crtc_state(). - Recheck if the port needs to be reset after all the atomic state is locked and async commits are waited on. v3: - Add intel_crtc_needs_link_reset(), instead of open-coding it, keep intel_crtc_has_encoders(). (Ville) - Fix state dumping and use a bitmask to track disabled CRTCs in intel_sanitize_all_crtcs(). (Ville) - Set internal in intel_atomic_state right after allocating it. (Ville) - Recheck all CRTCs (not yet force-disabled) after a CRTC is force-disabled for any reason (not only due to a link state) in intel_sanitize_all_crtcs(). - Reduce delay after CRTC disabling to 20ms, and use the simpler msleep(). - Clarify code comment about HPD behaviour in intel_sanitize_all_crtcs(). - Move all the TC link reset logic to intel_tc.c . - Cancel the link reset work synchronously during system suspend, driver unload and shutdown. v4: - Rebased on previous patch, which allows calling the TC port suspend/cleanup handlers without modeset locks held; remove the display driver suspended assert from the link reset work accordingly. v5: (Ville) - Remove reset work canceling from intel_ddi_pre_pll_enable(). - Track a crtc vs. pipe mask in intel_sanitize_all_crtcs(). - Add reset_link_commit() to clarify the intel_modeset_lock_ctx_retry loop. Cc: Kai-Heng Feng <kai.heng.feng@canonical.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Tested-by: Kai-Heng Feng <kai.heng.feng@canonical.com> Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/5860 Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230512195513.2699-2-imre.deak@intel.com
2023-05-12 22:55:13 +03:00
bool needs_link_reset;
if (crtc_state->hw.active) {
struct intel_plane *plane;
/* Disable everything but the primary plane */
for_each_intel_plane_on_crtc(display->drm, crtc, plane) {
const struct intel_plane_state *plane_state =
to_intel_plane_state(plane->base.state);
if (plane_state->uapi.visible &&
plane->base.type != DRM_PLANE_TYPE_PRIMARY)
intel_plane_disable_noatomic(crtc, plane);
}
/* Disable any background color/etc. set by the BIOS */
intel_color_commit_noarm(NULL, crtc_state);
intel_color_commit_arm(NULL, crtc_state);
}
drm/i915/tc: Reset TypeC PHYs left enabled in DP-alt mode after the sink disconnects If the output on a DP-alt link with its sink disconnected is kept enabled for too long (about 20 sec), then some IOM/TCSS firmware timeout will cause havoc on the PCI bus, at least for other GFX devices on it which will stop powering up. Since user space is not guaranteed to do a disabling modeset in time, switch such disconnected but active links to TBT mode - which is without such shortcomings - with a 2 second delay. If the above condition is detected already during the driver load/system resume sanitization step disable the output instead, as at that point no user space or kernel client depends on a consistent output state yet and because subsequent atomic modeset on such connectors - without the actual sink capabilities available - can fail. An active/disconnected port as above will also block the HPD status of other active/disconnected ports to get updated (stuck in the connected state), until the former port is disabled, its PHY is disconnected and a ~10 ms delay has elapsed. This means the link state for all TypeC ports/CRTCs must be rechecked after a CRTC is disabled due to the above reason. For this disconnect the PHY synchronously after the CRTC/port is disabled and recheck all CRTCs for the above condition whenever such a port is disabled. To account for a race condition during driver loading where the sink is disconnected after the above sanitization step and before the HPD interrupts get enabled, do an explicit check/link reset if needed from the encoder's late_register hook, which is called after the HPD interrupts are enabled already. v2: - Handle an active/disconnected port blocking the HPD state update of another active/disconnected port. - Cancel the delayed work resetting the link also from the encoder enable/suspend/shutdown hooks. - Rebase on the earlier intel_modeset_lock_ctx_retry() addition, fixing here the missed atomic state reset in case of a retry. - Fix handling of an error return from intel_atomic_get_crtc_state(). - Recheck if the port needs to be reset after all the atomic state is locked and async commits are waited on. v3: - Add intel_crtc_needs_link_reset(), instead of open-coding it, keep intel_crtc_has_encoders(). (Ville) - Fix state dumping and use a bitmask to track disabled CRTCs in intel_sanitize_all_crtcs(). (Ville) - Set internal in intel_atomic_state right after allocating it. (Ville) - Recheck all CRTCs (not yet force-disabled) after a CRTC is force-disabled for any reason (not only due to a link state) in intel_sanitize_all_crtcs(). - Reduce delay after CRTC disabling to 20ms, and use the simpler msleep(). - Clarify code comment about HPD behaviour in intel_sanitize_all_crtcs(). - Move all the TC link reset logic to intel_tc.c . - Cancel the link reset work synchronously during system suspend, driver unload and shutdown. v4: - Rebased on previous patch, which allows calling the TC port suspend/cleanup handlers without modeset locks held; remove the display driver suspended assert from the link reset work accordingly. v5: (Ville) - Remove reset work canceling from intel_ddi_pre_pll_enable(). - Track a crtc vs. pipe mask in intel_sanitize_all_crtcs(). - Add reset_link_commit() to clarify the intel_modeset_lock_ctx_retry loop. Cc: Kai-Heng Feng <kai.heng.feng@canonical.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Tested-by: Kai-Heng Feng <kai.heng.feng@canonical.com> Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/5860 Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230512195513.2699-2-imre.deak@intel.com
2023-05-12 22:55:13 +03:00
if (!crtc_state->hw.active ||
intel_crtc_is_joiner_secondary(crtc_state))
drm/i915/tc: Reset TypeC PHYs left enabled in DP-alt mode after the sink disconnects If the output on a DP-alt link with its sink disconnected is kept enabled for too long (about 20 sec), then some IOM/TCSS firmware timeout will cause havoc on the PCI bus, at least for other GFX devices on it which will stop powering up. Since user space is not guaranteed to do a disabling modeset in time, switch such disconnected but active links to TBT mode - which is without such shortcomings - with a 2 second delay. If the above condition is detected already during the driver load/system resume sanitization step disable the output instead, as at that point no user space or kernel client depends on a consistent output state yet and because subsequent atomic modeset on such connectors - without the actual sink capabilities available - can fail. An active/disconnected port as above will also block the HPD status of other active/disconnected ports to get updated (stuck in the connected state), until the former port is disabled, its PHY is disconnected and a ~10 ms delay has elapsed. This means the link state for all TypeC ports/CRTCs must be rechecked after a CRTC is disabled due to the above reason. For this disconnect the PHY synchronously after the CRTC/port is disabled and recheck all CRTCs for the above condition whenever such a port is disabled. To account for a race condition during driver loading where the sink is disconnected after the above sanitization step and before the HPD interrupts get enabled, do an explicit check/link reset if needed from the encoder's late_register hook, which is called after the HPD interrupts are enabled already. v2: - Handle an active/disconnected port blocking the HPD state update of another active/disconnected port. - Cancel the delayed work resetting the link also from the encoder enable/suspend/shutdown hooks. - Rebase on the earlier intel_modeset_lock_ctx_retry() addition, fixing here the missed atomic state reset in case of a retry. - Fix handling of an error return from intel_atomic_get_crtc_state(). - Recheck if the port needs to be reset after all the atomic state is locked and async commits are waited on. v3: - Add intel_crtc_needs_link_reset(), instead of open-coding it, keep intel_crtc_has_encoders(). (Ville) - Fix state dumping and use a bitmask to track disabled CRTCs in intel_sanitize_all_crtcs(). (Ville) - Set internal in intel_atomic_state right after allocating it. (Ville) - Recheck all CRTCs (not yet force-disabled) after a CRTC is force-disabled for any reason (not only due to a link state) in intel_sanitize_all_crtcs(). - Reduce delay after CRTC disabling to 20ms, and use the simpler msleep(). - Clarify code comment about HPD behaviour in intel_sanitize_all_crtcs(). - Move all the TC link reset logic to intel_tc.c . - Cancel the link reset work synchronously during system suspend, driver unload and shutdown. v4: - Rebased on previous patch, which allows calling the TC port suspend/cleanup handlers without modeset locks held; remove the display driver suspended assert from the link reset work accordingly. v5: (Ville) - Remove reset work canceling from intel_ddi_pre_pll_enable(). - Track a crtc vs. pipe mask in intel_sanitize_all_crtcs(). - Add reset_link_commit() to clarify the intel_modeset_lock_ctx_retry loop. Cc: Kai-Heng Feng <kai.heng.feng@canonical.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Tested-by: Kai-Heng Feng <kai.heng.feng@canonical.com> Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/5860 Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230512195513.2699-2-imre.deak@intel.com
2023-05-12 22:55:13 +03:00
return false;
needs_link_reset = intel_crtc_needs_link_reset(crtc);
/*
* Adjust the state of the output pipe according to whether we have
* active connectors/encoders.
*/
drm/i915/tc: Reset TypeC PHYs left enabled in DP-alt mode after the sink disconnects If the output on a DP-alt link with its sink disconnected is kept enabled for too long (about 20 sec), then some IOM/TCSS firmware timeout will cause havoc on the PCI bus, at least for other GFX devices on it which will stop powering up. Since user space is not guaranteed to do a disabling modeset in time, switch such disconnected but active links to TBT mode - which is without such shortcomings - with a 2 second delay. If the above condition is detected already during the driver load/system resume sanitization step disable the output instead, as at that point no user space or kernel client depends on a consistent output state yet and because subsequent atomic modeset on such connectors - without the actual sink capabilities available - can fail. An active/disconnected port as above will also block the HPD status of other active/disconnected ports to get updated (stuck in the connected state), until the former port is disabled, its PHY is disconnected and a ~10 ms delay has elapsed. This means the link state for all TypeC ports/CRTCs must be rechecked after a CRTC is disabled due to the above reason. For this disconnect the PHY synchronously after the CRTC/port is disabled and recheck all CRTCs for the above condition whenever such a port is disabled. To account for a race condition during driver loading where the sink is disconnected after the above sanitization step and before the HPD interrupts get enabled, do an explicit check/link reset if needed from the encoder's late_register hook, which is called after the HPD interrupts are enabled already. v2: - Handle an active/disconnected port blocking the HPD state update of another active/disconnected port. - Cancel the delayed work resetting the link also from the encoder enable/suspend/shutdown hooks. - Rebase on the earlier intel_modeset_lock_ctx_retry() addition, fixing here the missed atomic state reset in case of a retry. - Fix handling of an error return from intel_atomic_get_crtc_state(). - Recheck if the port needs to be reset after all the atomic state is locked and async commits are waited on. v3: - Add intel_crtc_needs_link_reset(), instead of open-coding it, keep intel_crtc_has_encoders(). (Ville) - Fix state dumping and use a bitmask to track disabled CRTCs in intel_sanitize_all_crtcs(). (Ville) - Set internal in intel_atomic_state right after allocating it. (Ville) - Recheck all CRTCs (not yet force-disabled) after a CRTC is force-disabled for any reason (not only due to a link state) in intel_sanitize_all_crtcs(). - Reduce delay after CRTC disabling to 20ms, and use the simpler msleep(). - Clarify code comment about HPD behaviour in intel_sanitize_all_crtcs(). - Move all the TC link reset logic to intel_tc.c . - Cancel the link reset work synchronously during system suspend, driver unload and shutdown. v4: - Rebased on previous patch, which allows calling the TC port suspend/cleanup handlers without modeset locks held; remove the display driver suspended assert from the link reset work accordingly. v5: (Ville) - Remove reset work canceling from intel_ddi_pre_pll_enable(). - Track a crtc vs. pipe mask in intel_sanitize_all_crtcs(). - Add reset_link_commit() to clarify the intel_modeset_lock_ctx_retry loop. Cc: Kai-Heng Feng <kai.heng.feng@canonical.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Tested-by: Kai-Heng Feng <kai.heng.feng@canonical.com> Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/5860 Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230512195513.2699-2-imre.deak@intel.com
2023-05-12 22:55:13 +03:00
if (!needs_link_reset && intel_crtc_has_encoders(crtc))
return false;
intel_crtc_disable_noatomic(crtc, ctx);
/*
* The HPD state on other active/disconnected TC ports may be stuck in
* the connected state until this port is disabled and a ~10ms delay has
* passed, wait here for that so that sanitizing other CRTCs will see the
* up-to-date HPD state.
*/
if (needs_link_reset)
msleep(20);
return true;
}
static void intel_sanitize_all_crtcs(struct intel_display *display,
drm/i915/tc: Reset TypeC PHYs left enabled in DP-alt mode after the sink disconnects If the output on a DP-alt link with its sink disconnected is kept enabled for too long (about 20 sec), then some IOM/TCSS firmware timeout will cause havoc on the PCI bus, at least for other GFX devices on it which will stop powering up. Since user space is not guaranteed to do a disabling modeset in time, switch such disconnected but active links to TBT mode - which is without such shortcomings - with a 2 second delay. If the above condition is detected already during the driver load/system resume sanitization step disable the output instead, as at that point no user space or kernel client depends on a consistent output state yet and because subsequent atomic modeset on such connectors - without the actual sink capabilities available - can fail. An active/disconnected port as above will also block the HPD status of other active/disconnected ports to get updated (stuck in the connected state), until the former port is disabled, its PHY is disconnected and a ~10 ms delay has elapsed. This means the link state for all TypeC ports/CRTCs must be rechecked after a CRTC is disabled due to the above reason. For this disconnect the PHY synchronously after the CRTC/port is disabled and recheck all CRTCs for the above condition whenever such a port is disabled. To account for a race condition during driver loading where the sink is disconnected after the above sanitization step and before the HPD interrupts get enabled, do an explicit check/link reset if needed from the encoder's late_register hook, which is called after the HPD interrupts are enabled already. v2: - Handle an active/disconnected port blocking the HPD state update of another active/disconnected port. - Cancel the delayed work resetting the link also from the encoder enable/suspend/shutdown hooks. - Rebase on the earlier intel_modeset_lock_ctx_retry() addition, fixing here the missed atomic state reset in case of a retry. - Fix handling of an error return from intel_atomic_get_crtc_state(). - Recheck if the port needs to be reset after all the atomic state is locked and async commits are waited on. v3: - Add intel_crtc_needs_link_reset(), instead of open-coding it, keep intel_crtc_has_encoders(). (Ville) - Fix state dumping and use a bitmask to track disabled CRTCs in intel_sanitize_all_crtcs(). (Ville) - Set internal in intel_atomic_state right after allocating it. (Ville) - Recheck all CRTCs (not yet force-disabled) after a CRTC is force-disabled for any reason (not only due to a link state) in intel_sanitize_all_crtcs(). - Reduce delay after CRTC disabling to 20ms, and use the simpler msleep(). - Clarify code comment about HPD behaviour in intel_sanitize_all_crtcs(). - Move all the TC link reset logic to intel_tc.c . - Cancel the link reset work synchronously during system suspend, driver unload and shutdown. v4: - Rebased on previous patch, which allows calling the TC port suspend/cleanup handlers without modeset locks held; remove the display driver suspended assert from the link reset work accordingly. v5: (Ville) - Remove reset work canceling from intel_ddi_pre_pll_enable(). - Track a crtc vs. pipe mask in intel_sanitize_all_crtcs(). - Add reset_link_commit() to clarify the intel_modeset_lock_ctx_retry loop. Cc: Kai-Heng Feng <kai.heng.feng@canonical.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Tested-by: Kai-Heng Feng <kai.heng.feng@canonical.com> Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/5860 Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230512195513.2699-2-imre.deak@intel.com
2023-05-12 22:55:13 +03:00
struct drm_modeset_acquire_ctx *ctx)
{
struct intel_crtc *crtc;
u32 crtcs_forced_off = 0;
/*
* An active and disconnected TypeC port prevents the HPD live state
* to get updated on other active/disconnected TypeC ports, so after
* a port gets disabled the CRTCs using other TypeC ports must be
* rechecked wrt. their link status.
*/
for (;;) {
u32 old_mask = crtcs_forced_off;
for_each_intel_crtc(display->drm, crtc) {
drm/i915/tc: Reset TypeC PHYs left enabled in DP-alt mode after the sink disconnects If the output on a DP-alt link with its sink disconnected is kept enabled for too long (about 20 sec), then some IOM/TCSS firmware timeout will cause havoc on the PCI bus, at least for other GFX devices on it which will stop powering up. Since user space is not guaranteed to do a disabling modeset in time, switch such disconnected but active links to TBT mode - which is without such shortcomings - with a 2 second delay. If the above condition is detected already during the driver load/system resume sanitization step disable the output instead, as at that point no user space or kernel client depends on a consistent output state yet and because subsequent atomic modeset on such connectors - without the actual sink capabilities available - can fail. An active/disconnected port as above will also block the HPD status of other active/disconnected ports to get updated (stuck in the connected state), until the former port is disabled, its PHY is disconnected and a ~10 ms delay has elapsed. This means the link state for all TypeC ports/CRTCs must be rechecked after a CRTC is disabled due to the above reason. For this disconnect the PHY synchronously after the CRTC/port is disabled and recheck all CRTCs for the above condition whenever such a port is disabled. To account for a race condition during driver loading where the sink is disconnected after the above sanitization step and before the HPD interrupts get enabled, do an explicit check/link reset if needed from the encoder's late_register hook, which is called after the HPD interrupts are enabled already. v2: - Handle an active/disconnected port blocking the HPD state update of another active/disconnected port. - Cancel the delayed work resetting the link also from the encoder enable/suspend/shutdown hooks. - Rebase on the earlier intel_modeset_lock_ctx_retry() addition, fixing here the missed atomic state reset in case of a retry. - Fix handling of an error return from intel_atomic_get_crtc_state(). - Recheck if the port needs to be reset after all the atomic state is locked and async commits are waited on. v3: - Add intel_crtc_needs_link_reset(), instead of open-coding it, keep intel_crtc_has_encoders(). (Ville) - Fix state dumping and use a bitmask to track disabled CRTCs in intel_sanitize_all_crtcs(). (Ville) - Set internal in intel_atomic_state right after allocating it. (Ville) - Recheck all CRTCs (not yet force-disabled) after a CRTC is force-disabled for any reason (not only due to a link state) in intel_sanitize_all_crtcs(). - Reduce delay after CRTC disabling to 20ms, and use the simpler msleep(). - Clarify code comment about HPD behaviour in intel_sanitize_all_crtcs(). - Move all the TC link reset logic to intel_tc.c . - Cancel the link reset work synchronously during system suspend, driver unload and shutdown. v4: - Rebased on previous patch, which allows calling the TC port suspend/cleanup handlers without modeset locks held; remove the display driver suspended assert from the link reset work accordingly. v5: (Ville) - Remove reset work canceling from intel_ddi_pre_pll_enable(). - Track a crtc vs. pipe mask in intel_sanitize_all_crtcs(). - Add reset_link_commit() to clarify the intel_modeset_lock_ctx_retry loop. Cc: Kai-Heng Feng <kai.heng.feng@canonical.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Tested-by: Kai-Heng Feng <kai.heng.feng@canonical.com> Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/5860 Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230512195513.2699-2-imre.deak@intel.com
2023-05-12 22:55:13 +03:00
u32 crtc_mask = drm_crtc_mask(&crtc->base);
if (crtcs_forced_off & crtc_mask)
continue;
if (intel_sanitize_crtc(crtc, ctx))
crtcs_forced_off |= crtc_mask;
}
if (crtcs_forced_off == old_mask)
break;
}
for_each_intel_crtc(display->drm, crtc) {
drm/i915/tc: Reset TypeC PHYs left enabled in DP-alt mode after the sink disconnects If the output on a DP-alt link with its sink disconnected is kept enabled for too long (about 20 sec), then some IOM/TCSS firmware timeout will cause havoc on the PCI bus, at least for other GFX devices on it which will stop powering up. Since user space is not guaranteed to do a disabling modeset in time, switch such disconnected but active links to TBT mode - which is without such shortcomings - with a 2 second delay. If the above condition is detected already during the driver load/system resume sanitization step disable the output instead, as at that point no user space or kernel client depends on a consistent output state yet and because subsequent atomic modeset on such connectors - without the actual sink capabilities available - can fail. An active/disconnected port as above will also block the HPD status of other active/disconnected ports to get updated (stuck in the connected state), until the former port is disabled, its PHY is disconnected and a ~10 ms delay has elapsed. This means the link state for all TypeC ports/CRTCs must be rechecked after a CRTC is disabled due to the above reason. For this disconnect the PHY synchronously after the CRTC/port is disabled and recheck all CRTCs for the above condition whenever such a port is disabled. To account for a race condition during driver loading where the sink is disconnected after the above sanitization step and before the HPD interrupts get enabled, do an explicit check/link reset if needed from the encoder's late_register hook, which is called after the HPD interrupts are enabled already. v2: - Handle an active/disconnected port blocking the HPD state update of another active/disconnected port. - Cancel the delayed work resetting the link also from the encoder enable/suspend/shutdown hooks. - Rebase on the earlier intel_modeset_lock_ctx_retry() addition, fixing here the missed atomic state reset in case of a retry. - Fix handling of an error return from intel_atomic_get_crtc_state(). - Recheck if the port needs to be reset after all the atomic state is locked and async commits are waited on. v3: - Add intel_crtc_needs_link_reset(), instead of open-coding it, keep intel_crtc_has_encoders(). (Ville) - Fix state dumping and use a bitmask to track disabled CRTCs in intel_sanitize_all_crtcs(). (Ville) - Set internal in intel_atomic_state right after allocating it. (Ville) - Recheck all CRTCs (not yet force-disabled) after a CRTC is force-disabled for any reason (not only due to a link state) in intel_sanitize_all_crtcs(). - Reduce delay after CRTC disabling to 20ms, and use the simpler msleep(). - Clarify code comment about HPD behaviour in intel_sanitize_all_crtcs(). - Move all the TC link reset logic to intel_tc.c . - Cancel the link reset work synchronously during system suspend, driver unload and shutdown. v4: - Rebased on previous patch, which allows calling the TC port suspend/cleanup handlers without modeset locks held; remove the display driver suspended assert from the link reset work accordingly. v5: (Ville) - Remove reset work canceling from intel_ddi_pre_pll_enable(). - Track a crtc vs. pipe mask in intel_sanitize_all_crtcs(). - Add reset_link_commit() to clarify the intel_modeset_lock_ctx_retry loop. Cc: Kai-Heng Feng <kai.heng.feng@canonical.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Tested-by: Kai-Heng Feng <kai.heng.feng@canonical.com> Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/5860 Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230512195513.2699-2-imre.deak@intel.com
2023-05-12 22:55:13 +03:00
struct intel_crtc_state *crtc_state =
to_intel_crtc_state(crtc->base.state);
intel_crtc_state_dump(crtc_state, NULL, "setup_hw_state");
}
}
static bool has_bogus_dpll_config(const struct intel_crtc_state *crtc_state)
{
struct intel_display *display = to_intel_display(crtc_state);
/*
* Some SNB BIOSen (eg. ASUS K53SV) are known to misprogram
* the hardware when a high res displays plugged in. DPLL P
* divider is zero, and the pipe timings are bonkers. We'll
* try to disable everything in that case.
*
* FIXME would be nice to be able to sanitize this state
* without several WARNs, but for now let's take the easy
* road.
*/
return display->platform.sandybridge &&
crtc_state->hw.active &&
crtc_state->intel_dpll &&
crtc_state->port_clock == 0;
}
static void intel_sanitize_encoder(struct intel_encoder *encoder)
{
struct intel_display *display = to_intel_display(encoder);
struct intel_connector *connector;
struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc);
struct intel_crtc_state *crtc_state = crtc ?
to_intel_crtc_state(crtc->base.state) : NULL;
drm/i915/mtl: Add support for PM DEMAND MTL introduces a new way to instruct the PUnit with power and bandwidth requirements of DE. Add the functionality to program the registers and handle waits using interrupts. The current wait time for timeouts is programmed for 10 msecs to factor in the worst case scenarios. Changes made to use REG_BIT for a register that we touched(GEN8_DE_MISC_IER _MMIO). Wa_14016740474 is added which applies to Xe_LPD+ display v2: checkpatch warning fixes, simplify program pmdemand part v3: update to dbufs and pipes values to pmdemand register(stan) Removed the macro usage in update_pmdemand_values() v4: move the pmdemand_pre_plane_update before cdclk update pmdemand_needs_update included cdclk params comparisons pmdemand_state NULL check (Gustavo) pmdemand.o in sorted order in the makefile (Jani) update pmdemand misc irq handler loop (Gustavo) active phys bitmask and programming correction (Gustavo) v5: simplify pmdemand_state structure simplify methods to find active phys and max port clock Timeout in case of previou pmdemand task pending (Gustavo) v6: rebasing updates to max_ddiclk calculations (Gustavo) updates to active_phys count method (Gustavo) v7: use two separate loop to iterate throug old and new crtc states to calculate the active phys (Gustavo) v8: use uniform function names (Gustavo) v9: For phys change iterate through connectors (Imre) Look for change in phys for pmdemand update (Gustavo, Imre) Some more stlying changes (Imre) Update pmdemand state during HW readout/sanitize (Imre) v10: Fix CI checkpatch warnings v11: use correct pmdemand object pointer during hw readout, simplify the check for phys need update (Gustavo) v12: Handle possible non serialize cases (Imre) Initialise also pmdemand params HW readout (Imre) Update active phys mask during sanitize calls (Imre) Check TC/encoder changes to limit connector update (Imre) v13: Check display version before accessing pmdemand functions v14: Move is_serialized to intel_global_state.c simplify update params and other stlying issues (Imre) Bspec: 66451, 64636, 64602, 64603 Cc: Matt Atwood <matthew.s.atwood@intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: José Roberto de Souza <jose.souza@intel.com> Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: Mika Kahola <mika.kahola@intel.com> Signed-off-by: Vinod Govindapillai <vinod.govindapillai@intel.com> Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> #v4 Acked-by: Gustavo Sousa <gustavo.sousa@intel.com> #v11 Reviewed-by: Imre Deak <imre.deak@intel.com> [RK: Fixed minor typo in one of the comments. s/qclck_gc/qclk_gv/] Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230606201032.347449-1-vinod.govindapillai@intel.com
2023-06-06 23:10:32 +03:00
struct intel_pmdemand_state *pmdemand_state =
to_intel_pmdemand_state(display->pmdemand.obj.state);
/*
* We need to check both for a crtc link (meaning that the encoder is
* active and trying to read from a pipe) and the pipe itself being
* active.
*/
bool has_active_crtc = crtc_state &&
crtc_state->hw.active;
if (crtc_state && has_bogus_dpll_config(crtc_state)) {
drm_dbg_kms(display->drm,
"BIOS has misprogrammed the hardware. Disabling pipe %c\n",
pipe_name(crtc->pipe));
has_active_crtc = false;
}
connector = intel_encoder_find_connector(encoder);
if (connector && !has_active_crtc) {
drm_dbg_kms(display->drm,
"[ENCODER:%d:%s] has active connectors but no active pipe!\n",
encoder->base.base.id,
encoder->base.name);
drm/i915/mtl: Add support for PM DEMAND MTL introduces a new way to instruct the PUnit with power and bandwidth requirements of DE. Add the functionality to program the registers and handle waits using interrupts. The current wait time for timeouts is programmed for 10 msecs to factor in the worst case scenarios. Changes made to use REG_BIT for a register that we touched(GEN8_DE_MISC_IER _MMIO). Wa_14016740474 is added which applies to Xe_LPD+ display v2: checkpatch warning fixes, simplify program pmdemand part v3: update to dbufs and pipes values to pmdemand register(stan) Removed the macro usage in update_pmdemand_values() v4: move the pmdemand_pre_plane_update before cdclk update pmdemand_needs_update included cdclk params comparisons pmdemand_state NULL check (Gustavo) pmdemand.o in sorted order in the makefile (Jani) update pmdemand misc irq handler loop (Gustavo) active phys bitmask and programming correction (Gustavo) v5: simplify pmdemand_state structure simplify methods to find active phys and max port clock Timeout in case of previou pmdemand task pending (Gustavo) v6: rebasing updates to max_ddiclk calculations (Gustavo) updates to active_phys count method (Gustavo) v7: use two separate loop to iterate throug old and new crtc states to calculate the active phys (Gustavo) v8: use uniform function names (Gustavo) v9: For phys change iterate through connectors (Imre) Look for change in phys for pmdemand update (Gustavo, Imre) Some more stlying changes (Imre) Update pmdemand state during HW readout/sanitize (Imre) v10: Fix CI checkpatch warnings v11: use correct pmdemand object pointer during hw readout, simplify the check for phys need update (Gustavo) v12: Handle possible non serialize cases (Imre) Initialise also pmdemand params HW readout (Imre) Update active phys mask during sanitize calls (Imre) Check TC/encoder changes to limit connector update (Imre) v13: Check display version before accessing pmdemand functions v14: Move is_serialized to intel_global_state.c simplify update params and other stlying issues (Imre) Bspec: 66451, 64636, 64602, 64603 Cc: Matt Atwood <matthew.s.atwood@intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: José Roberto de Souza <jose.souza@intel.com> Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: Mika Kahola <mika.kahola@intel.com> Signed-off-by: Vinod Govindapillai <vinod.govindapillai@intel.com> Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> #v4 Acked-by: Gustavo Sousa <gustavo.sousa@intel.com> #v11 Reviewed-by: Imre Deak <imre.deak@intel.com> [RK: Fixed minor typo in one of the comments. s/qclck_gc/qclk_gv/] Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230606201032.347449-1-vinod.govindapillai@intel.com
2023-06-06 23:10:32 +03:00
/* Clear the corresponding bit in pmdemand active phys mask */
intel_pmdemand_update_phys_mask(display, encoder,
drm/i915/mtl: Add support for PM DEMAND MTL introduces a new way to instruct the PUnit with power and bandwidth requirements of DE. Add the functionality to program the registers and handle waits using interrupts. The current wait time for timeouts is programmed for 10 msecs to factor in the worst case scenarios. Changes made to use REG_BIT for a register that we touched(GEN8_DE_MISC_IER _MMIO). Wa_14016740474 is added which applies to Xe_LPD+ display v2: checkpatch warning fixes, simplify program pmdemand part v3: update to dbufs and pipes values to pmdemand register(stan) Removed the macro usage in update_pmdemand_values() v4: move the pmdemand_pre_plane_update before cdclk update pmdemand_needs_update included cdclk params comparisons pmdemand_state NULL check (Gustavo) pmdemand.o in sorted order in the makefile (Jani) update pmdemand misc irq handler loop (Gustavo) active phys bitmask and programming correction (Gustavo) v5: simplify pmdemand_state structure simplify methods to find active phys and max port clock Timeout in case of previou pmdemand task pending (Gustavo) v6: rebasing updates to max_ddiclk calculations (Gustavo) updates to active_phys count method (Gustavo) v7: use two separate loop to iterate throug old and new crtc states to calculate the active phys (Gustavo) v8: use uniform function names (Gustavo) v9: For phys change iterate through connectors (Imre) Look for change in phys for pmdemand update (Gustavo, Imre) Some more stlying changes (Imre) Update pmdemand state during HW readout/sanitize (Imre) v10: Fix CI checkpatch warnings v11: use correct pmdemand object pointer during hw readout, simplify the check for phys need update (Gustavo) v12: Handle possible non serialize cases (Imre) Initialise also pmdemand params HW readout (Imre) Update active phys mask during sanitize calls (Imre) Check TC/encoder changes to limit connector update (Imre) v13: Check display version before accessing pmdemand functions v14: Move is_serialized to intel_global_state.c simplify update params and other stlying issues (Imre) Bspec: 66451, 64636, 64602, 64603 Cc: Matt Atwood <matthew.s.atwood@intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: José Roberto de Souza <jose.souza@intel.com> Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: Mika Kahola <mika.kahola@intel.com> Signed-off-by: Vinod Govindapillai <vinod.govindapillai@intel.com> Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> #v4 Acked-by: Gustavo Sousa <gustavo.sousa@intel.com> #v11 Reviewed-by: Imre Deak <imre.deak@intel.com> [RK: Fixed minor typo in one of the comments. s/qclck_gc/qclk_gv/] Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230606201032.347449-1-vinod.govindapillai@intel.com
2023-06-06 23:10:32 +03:00
pmdemand_state, false);
/*
* Connector is active, but has no active pipe. This is fallout
* from our resume register restoring. Disable the encoder
* manually again.
*/
if (crtc_state) {
struct drm_encoder *best_encoder;
drm_dbg_kms(display->drm,
"[ENCODER:%d:%s] manually disabled\n",
encoder->base.base.id,
encoder->base.name);
/* avoid oopsing in case the hooks consult best_encoder */
best_encoder = connector->base.state->best_encoder;
connector->base.state->best_encoder = &encoder->base;
/* FIXME NULL atomic state passed! */
if (encoder->disable)
encoder->disable(NULL, encoder, crtc_state,
connector->base.state);
if (encoder->post_disable)
encoder->post_disable(NULL, encoder, crtc_state,
connector->base.state);
connector->base.state->best_encoder = best_encoder;
}
encoder->base.crtc = NULL;
/*
* Inconsistent output/port/pipe state happens presumably due to
* a bug in one of the get_hw_state functions. Or someplace else
* in our code, like the register restore mess on resume. Clamp
* things to off as a safer default.
*/
connector->base.dpms = DRM_MODE_DPMS_OFF;
connector->base.encoder = NULL;
}
/* notify opregion of the sanitized encoder state */
intel_opregion_notify_encoder(encoder, connector && has_active_crtc);
if (HAS_DDI(display))
intel_ddi_sanitize_encoder_pll_mapping(encoder);
}
/* FIXME read out full plane state for all planes */
static void readout_plane_state(struct intel_display *display)
{
struct intel_plane *plane;
struct intel_crtc *crtc;
for_each_intel_plane(display->drm, plane) {
struct intel_plane_state *plane_state =
to_intel_plane_state(plane->base.state);
struct intel_crtc_state *crtc_state;
enum pipe pipe = PIPE_A;
bool visible;
visible = plane->get_hw_state(plane, &pipe);
crtc = intel_crtc_for_pipe(display, pipe);
crtc_state = to_intel_crtc_state(crtc->base.state);
intel_set_plane_visible(crtc_state, plane_state, visible);
drm_dbg_kms(display->drm,
"[PLANE:%d:%s] hw state readout: %s, pipe %c\n",
plane->base.base.id, plane->base.name,
str_enabled_disabled(visible), pipe_name(pipe));
}
for_each_intel_crtc(display->drm, crtc) {
struct intel_crtc_state *crtc_state =
to_intel_crtc_state(crtc->base.state);
intel_plane_fixup_bitmasks(crtc_state);
}
}
static void intel_modeset_readout_hw_state(struct intel_display *display)
{
drm/i915/mtl: Add support for PM DEMAND MTL introduces a new way to instruct the PUnit with power and bandwidth requirements of DE. Add the functionality to program the registers and handle waits using interrupts. The current wait time for timeouts is programmed for 10 msecs to factor in the worst case scenarios. Changes made to use REG_BIT for a register that we touched(GEN8_DE_MISC_IER _MMIO). Wa_14016740474 is added which applies to Xe_LPD+ display v2: checkpatch warning fixes, simplify program pmdemand part v3: update to dbufs and pipes values to pmdemand register(stan) Removed the macro usage in update_pmdemand_values() v4: move the pmdemand_pre_plane_update before cdclk update pmdemand_needs_update included cdclk params comparisons pmdemand_state NULL check (Gustavo) pmdemand.o in sorted order in the makefile (Jani) update pmdemand misc irq handler loop (Gustavo) active phys bitmask and programming correction (Gustavo) v5: simplify pmdemand_state structure simplify methods to find active phys and max port clock Timeout in case of previou pmdemand task pending (Gustavo) v6: rebasing updates to max_ddiclk calculations (Gustavo) updates to active_phys count method (Gustavo) v7: use two separate loop to iterate throug old and new crtc states to calculate the active phys (Gustavo) v8: use uniform function names (Gustavo) v9: For phys change iterate through connectors (Imre) Look for change in phys for pmdemand update (Gustavo, Imre) Some more stlying changes (Imre) Update pmdemand state during HW readout/sanitize (Imre) v10: Fix CI checkpatch warnings v11: use correct pmdemand object pointer during hw readout, simplify the check for phys need update (Gustavo) v12: Handle possible non serialize cases (Imre) Initialise also pmdemand params HW readout (Imre) Update active phys mask during sanitize calls (Imre) Check TC/encoder changes to limit connector update (Imre) v13: Check display version before accessing pmdemand functions v14: Move is_serialized to intel_global_state.c simplify update params and other stlying issues (Imre) Bspec: 66451, 64636, 64602, 64603 Cc: Matt Atwood <matthew.s.atwood@intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: José Roberto de Souza <jose.souza@intel.com> Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: Mika Kahola <mika.kahola@intel.com> Signed-off-by: Vinod Govindapillai <vinod.govindapillai@intel.com> Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> #v4 Acked-by: Gustavo Sousa <gustavo.sousa@intel.com> #v11 Reviewed-by: Imre Deak <imre.deak@intel.com> [RK: Fixed minor typo in one of the comments. s/qclck_gc/qclk_gv/] Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230606201032.347449-1-vinod.govindapillai@intel.com
2023-06-06 23:10:32 +03:00
struct intel_pmdemand_state *pmdemand_state =
to_intel_pmdemand_state(display->pmdemand.obj.state);
enum pipe pipe;
struct intel_crtc *crtc;
struct intel_encoder *encoder;
struct intel_connector *connector;
struct drm_connector_list_iter conn_iter;
for_each_intel_crtc(display->drm, crtc) {
struct intel_crtc_state *crtc_state =
to_intel_crtc_state(crtc->base.state);
__drm_atomic_helper_crtc_destroy_state(&crtc_state->uapi);
intel_crtc_free_hw_state(crtc_state);
intel_crtc_state_reset(crtc_state, crtc);
intel_crtc_get_pipe_config(crtc_state);
crtc_state->hw.enable = crtc_state->hw.active;
crtc->base.enabled = crtc_state->hw.enable;
crtc->active = crtc_state->hw.active;
drm_dbg_kms(display->drm,
"[CRTC:%d:%s] hw state readout: %s\n",
crtc->base.base.id, crtc->base.name,
str_enabled_disabled(crtc_state->hw.active));
}
readout_plane_state(display);
for_each_intel_encoder(display->drm, encoder) {
struct intel_crtc_state *crtc_state = NULL;
pipe = 0;
if (encoder->get_hw_state(encoder, &pipe)) {
crtc = intel_crtc_for_pipe(display, pipe);
crtc_state = to_intel_crtc_state(crtc->base.state);
encoder->base.crtc = &crtc->base;
intel_encoder_get_config(encoder, crtc_state);
/* read out to secondary crtc as well for joiner */
if (crtc_state->joiner_pipes) {
struct intel_crtc *secondary_crtc;
/* encoder should read be linked to joiner primary */
WARN_ON(intel_crtc_is_joiner_secondary(crtc_state));
for_each_intel_crtc_in_pipe_mask(display->drm, secondary_crtc,
intel_crtc_joiner_secondary_pipes(crtc_state)) {
struct intel_crtc_state *secondary_crtc_state;
secondary_crtc_state = to_intel_crtc_state(secondary_crtc->base.state);
intel_encoder_get_config(encoder, secondary_crtc_state);
}
}
drm/i915/mtl: Add support for PM DEMAND MTL introduces a new way to instruct the PUnit with power and bandwidth requirements of DE. Add the functionality to program the registers and handle waits using interrupts. The current wait time for timeouts is programmed for 10 msecs to factor in the worst case scenarios. Changes made to use REG_BIT for a register that we touched(GEN8_DE_MISC_IER _MMIO). Wa_14016740474 is added which applies to Xe_LPD+ display v2: checkpatch warning fixes, simplify program pmdemand part v3: update to dbufs and pipes values to pmdemand register(stan) Removed the macro usage in update_pmdemand_values() v4: move the pmdemand_pre_plane_update before cdclk update pmdemand_needs_update included cdclk params comparisons pmdemand_state NULL check (Gustavo) pmdemand.o in sorted order in the makefile (Jani) update pmdemand misc irq handler loop (Gustavo) active phys bitmask and programming correction (Gustavo) v5: simplify pmdemand_state structure simplify methods to find active phys and max port clock Timeout in case of previou pmdemand task pending (Gustavo) v6: rebasing updates to max_ddiclk calculations (Gustavo) updates to active_phys count method (Gustavo) v7: use two separate loop to iterate throug old and new crtc states to calculate the active phys (Gustavo) v8: use uniform function names (Gustavo) v9: For phys change iterate through connectors (Imre) Look for change in phys for pmdemand update (Gustavo, Imre) Some more stlying changes (Imre) Update pmdemand state during HW readout/sanitize (Imre) v10: Fix CI checkpatch warnings v11: use correct pmdemand object pointer during hw readout, simplify the check for phys need update (Gustavo) v12: Handle possible non serialize cases (Imre) Initialise also pmdemand params HW readout (Imre) Update active phys mask during sanitize calls (Imre) Check TC/encoder changes to limit connector update (Imre) v13: Check display version before accessing pmdemand functions v14: Move is_serialized to intel_global_state.c simplify update params and other stlying issues (Imre) Bspec: 66451, 64636, 64602, 64603 Cc: Matt Atwood <matthew.s.atwood@intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: José Roberto de Souza <jose.souza@intel.com> Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: Mika Kahola <mika.kahola@intel.com> Signed-off-by: Vinod Govindapillai <vinod.govindapillai@intel.com> Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> #v4 Acked-by: Gustavo Sousa <gustavo.sousa@intel.com> #v11 Reviewed-by: Imre Deak <imre.deak@intel.com> [RK: Fixed minor typo in one of the comments. s/qclck_gc/qclk_gv/] Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230606201032.347449-1-vinod.govindapillai@intel.com
2023-06-06 23:10:32 +03:00
intel_pmdemand_update_phys_mask(display, encoder,
drm/i915/mtl: Add support for PM DEMAND MTL introduces a new way to instruct the PUnit with power and bandwidth requirements of DE. Add the functionality to program the registers and handle waits using interrupts. The current wait time for timeouts is programmed for 10 msecs to factor in the worst case scenarios. Changes made to use REG_BIT for a register that we touched(GEN8_DE_MISC_IER _MMIO). Wa_14016740474 is added which applies to Xe_LPD+ display v2: checkpatch warning fixes, simplify program pmdemand part v3: update to dbufs and pipes values to pmdemand register(stan) Removed the macro usage in update_pmdemand_values() v4: move the pmdemand_pre_plane_update before cdclk update pmdemand_needs_update included cdclk params comparisons pmdemand_state NULL check (Gustavo) pmdemand.o in sorted order in the makefile (Jani) update pmdemand misc irq handler loop (Gustavo) active phys bitmask and programming correction (Gustavo) v5: simplify pmdemand_state structure simplify methods to find active phys and max port clock Timeout in case of previou pmdemand task pending (Gustavo) v6: rebasing updates to max_ddiclk calculations (Gustavo) updates to active_phys count method (Gustavo) v7: use two separate loop to iterate throug old and new crtc states to calculate the active phys (Gustavo) v8: use uniform function names (Gustavo) v9: For phys change iterate through connectors (Imre) Look for change in phys for pmdemand update (Gustavo, Imre) Some more stlying changes (Imre) Update pmdemand state during HW readout/sanitize (Imre) v10: Fix CI checkpatch warnings v11: use correct pmdemand object pointer during hw readout, simplify the check for phys need update (Gustavo) v12: Handle possible non serialize cases (Imre) Initialise also pmdemand params HW readout (Imre) Update active phys mask during sanitize calls (Imre) Check TC/encoder changes to limit connector update (Imre) v13: Check display version before accessing pmdemand functions v14: Move is_serialized to intel_global_state.c simplify update params and other stlying issues (Imre) Bspec: 66451, 64636, 64602, 64603 Cc: Matt Atwood <matthew.s.atwood@intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: José Roberto de Souza <jose.souza@intel.com> Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: Mika Kahola <mika.kahola@intel.com> Signed-off-by: Vinod Govindapillai <vinod.govindapillai@intel.com> Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> #v4 Acked-by: Gustavo Sousa <gustavo.sousa@intel.com> #v11 Reviewed-by: Imre Deak <imre.deak@intel.com> [RK: Fixed minor typo in one of the comments. s/qclck_gc/qclk_gv/] Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230606201032.347449-1-vinod.govindapillai@intel.com
2023-06-06 23:10:32 +03:00
pmdemand_state,
true);
} else {
intel_pmdemand_update_phys_mask(display, encoder,
drm/i915/mtl: Add support for PM DEMAND MTL introduces a new way to instruct the PUnit with power and bandwidth requirements of DE. Add the functionality to program the registers and handle waits using interrupts. The current wait time for timeouts is programmed for 10 msecs to factor in the worst case scenarios. Changes made to use REG_BIT for a register that we touched(GEN8_DE_MISC_IER _MMIO). Wa_14016740474 is added which applies to Xe_LPD+ display v2: checkpatch warning fixes, simplify program pmdemand part v3: update to dbufs and pipes values to pmdemand register(stan) Removed the macro usage in update_pmdemand_values() v4: move the pmdemand_pre_plane_update before cdclk update pmdemand_needs_update included cdclk params comparisons pmdemand_state NULL check (Gustavo) pmdemand.o in sorted order in the makefile (Jani) update pmdemand misc irq handler loop (Gustavo) active phys bitmask and programming correction (Gustavo) v5: simplify pmdemand_state structure simplify methods to find active phys and max port clock Timeout in case of previou pmdemand task pending (Gustavo) v6: rebasing updates to max_ddiclk calculations (Gustavo) updates to active_phys count method (Gustavo) v7: use two separate loop to iterate throug old and new crtc states to calculate the active phys (Gustavo) v8: use uniform function names (Gustavo) v9: For phys change iterate through connectors (Imre) Look for change in phys for pmdemand update (Gustavo, Imre) Some more stlying changes (Imre) Update pmdemand state during HW readout/sanitize (Imre) v10: Fix CI checkpatch warnings v11: use correct pmdemand object pointer during hw readout, simplify the check for phys need update (Gustavo) v12: Handle possible non serialize cases (Imre) Initialise also pmdemand params HW readout (Imre) Update active phys mask during sanitize calls (Imre) Check TC/encoder changes to limit connector update (Imre) v13: Check display version before accessing pmdemand functions v14: Move is_serialized to intel_global_state.c simplify update params and other stlying issues (Imre) Bspec: 66451, 64636, 64602, 64603 Cc: Matt Atwood <matthew.s.atwood@intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: José Roberto de Souza <jose.souza@intel.com> Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: Mika Kahola <mika.kahola@intel.com> Signed-off-by: Vinod Govindapillai <vinod.govindapillai@intel.com> Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> #v4 Acked-by: Gustavo Sousa <gustavo.sousa@intel.com> #v11 Reviewed-by: Imre Deak <imre.deak@intel.com> [RK: Fixed minor typo in one of the comments. s/qclck_gc/qclk_gv/] Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230606201032.347449-1-vinod.govindapillai@intel.com
2023-06-06 23:10:32 +03:00
pmdemand_state,
false);
encoder->base.crtc = NULL;
}
if (encoder->sync_state)
encoder->sync_state(encoder, crtc_state);
drm_dbg_kms(display->drm,
"[ENCODER:%d:%s] hw state readout: %s, pipe %c\n",
encoder->base.base.id, encoder->base.name,
str_enabled_disabled(encoder->base.crtc),
pipe_name(pipe));
}
intel_dpll_readout_hw_state(display);
drm_connector_list_iter_begin(display->drm, &conn_iter);
for_each_intel_connector_iter(connector, &conn_iter) {
struct intel_crtc_state *crtc_state = NULL;
if (connector->get_hw_state(connector)) {
struct intel_crtc *crtc;
connector->base.dpms = DRM_MODE_DPMS_ON;
encoder = intel_attached_encoder(connector);
connector->base.encoder = &encoder->base;
crtc = to_intel_crtc(encoder->base.crtc);
crtc_state = crtc ? to_intel_crtc_state(crtc->base.state) : NULL;
if (crtc_state && crtc_state->hw.active) {
/*
* This has to be done during hardware readout
* because anything calling .crtc_disable may
* rely on the connector_mask being accurate.
*/
crtc_state->uapi.connector_mask |=
drm_connector_mask(&connector->base);
crtc_state->uapi.encoder_mask |=
drm_encoder_mask(&encoder->base);
}
} else {
connector->base.dpms = DRM_MODE_DPMS_OFF;
connector->base.encoder = NULL;
}
if (connector->sync_state)
connector->sync_state(connector, crtc_state);
drm_dbg_kms(display->drm,
"[CONNECTOR:%d:%s] hw state readout: %s\n",
connector->base.base.id, connector->base.name,
str_enabled_disabled(connector->base.encoder));
}
drm_connector_list_iter_end(&conn_iter);
for_each_intel_crtc(display->drm, crtc) {
struct intel_crtc_state *crtc_state =
to_intel_crtc_state(crtc->base.state);
struct intel_plane *plane;
/*
* The initial mode needs to be set in order to keep
* the atomic core happy. It wants a valid mode if the
* crtc's enabled, so we do the above call.
*
* But we don't set all the derived state fully, hence
* set a flag to indicate that a full recalculation is
* needed on the next commit.
*/
crtc_state->inherited = true;
if (crtc_state->hw.active) {
intel_crtc_update_active_timings(crtc_state,
crtc_state->vrr.enable);
intel_crtc_copy_hw_to_uapi_state(crtc_state);
}
for_each_intel_plane_on_crtc(display->drm, crtc, plane) {
const struct intel_plane_state *plane_state =
to_intel_plane_state(plane->base.state);
/*
* FIXME don't have the fb yet, so can't
* use intel_plane_data_rate() :(
*/
if (plane_state->uapi.visible)
crtc_state->data_rate[plane->id] =
4 * crtc_state->pixel_rate;
/*
* FIXME don't have the fb yet, so can't
* use plane->min_cdclk() :(
*/
if (plane_state->uapi.visible && plane->min_cdclk) {
if (crtc_state->double_wide || DISPLAY_VER(display) >= 10)
crtc_state->min_cdclk[plane->id] =
DIV_ROUND_UP(crtc_state->pixel_rate, 2);
else
crtc_state->min_cdclk[plane->id] =
crtc_state->pixel_rate;
}
drm_dbg_kms(display->drm,
"[PLANE:%d:%s] min_cdclk %d kHz\n",
plane->base.base.id, plane->base.name,
crtc_state->min_cdclk[plane->id]);
}
intel_pmdemand_update_port_clock(display, pmdemand_state, pipe,
drm/i915/mtl: Add support for PM DEMAND MTL introduces a new way to instruct the PUnit with power and bandwidth requirements of DE. Add the functionality to program the registers and handle waits using interrupts. The current wait time for timeouts is programmed for 10 msecs to factor in the worst case scenarios. Changes made to use REG_BIT for a register that we touched(GEN8_DE_MISC_IER _MMIO). Wa_14016740474 is added which applies to Xe_LPD+ display v2: checkpatch warning fixes, simplify program pmdemand part v3: update to dbufs and pipes values to pmdemand register(stan) Removed the macro usage in update_pmdemand_values() v4: move the pmdemand_pre_plane_update before cdclk update pmdemand_needs_update included cdclk params comparisons pmdemand_state NULL check (Gustavo) pmdemand.o in sorted order in the makefile (Jani) update pmdemand misc irq handler loop (Gustavo) active phys bitmask and programming correction (Gustavo) v5: simplify pmdemand_state structure simplify methods to find active phys and max port clock Timeout in case of previou pmdemand task pending (Gustavo) v6: rebasing updates to max_ddiclk calculations (Gustavo) updates to active_phys count method (Gustavo) v7: use two separate loop to iterate throug old and new crtc states to calculate the active phys (Gustavo) v8: use uniform function names (Gustavo) v9: For phys change iterate through connectors (Imre) Look for change in phys for pmdemand update (Gustavo, Imre) Some more stlying changes (Imre) Update pmdemand state during HW readout/sanitize (Imre) v10: Fix CI checkpatch warnings v11: use correct pmdemand object pointer during hw readout, simplify the check for phys need update (Gustavo) v12: Handle possible non serialize cases (Imre) Initialise also pmdemand params HW readout (Imre) Update active phys mask during sanitize calls (Imre) Check TC/encoder changes to limit connector update (Imre) v13: Check display version before accessing pmdemand functions v14: Move is_serialized to intel_global_state.c simplify update params and other stlying issues (Imre) Bspec: 66451, 64636, 64602, 64603 Cc: Matt Atwood <matthew.s.atwood@intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: José Roberto de Souza <jose.souza@intel.com> Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: Mika Kahola <mika.kahola@intel.com> Signed-off-by: Vinod Govindapillai <vinod.govindapillai@intel.com> Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> #v4 Acked-by: Gustavo Sousa <gustavo.sousa@intel.com> #v11 Reviewed-by: Imre Deak <imre.deak@intel.com> [RK: Fixed minor typo in one of the comments. s/qclck_gc/qclk_gv/] Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230606201032.347449-1-vinod.govindapillai@intel.com
2023-06-06 23:10:32 +03:00
crtc_state->port_clock);
}
drm/i915/mtl: Add support for PM DEMAND MTL introduces a new way to instruct the PUnit with power and bandwidth requirements of DE. Add the functionality to program the registers and handle waits using interrupts. The current wait time for timeouts is programmed for 10 msecs to factor in the worst case scenarios. Changes made to use REG_BIT for a register that we touched(GEN8_DE_MISC_IER _MMIO). Wa_14016740474 is added which applies to Xe_LPD+ display v2: checkpatch warning fixes, simplify program pmdemand part v3: update to dbufs and pipes values to pmdemand register(stan) Removed the macro usage in update_pmdemand_values() v4: move the pmdemand_pre_plane_update before cdclk update pmdemand_needs_update included cdclk params comparisons pmdemand_state NULL check (Gustavo) pmdemand.o in sorted order in the makefile (Jani) update pmdemand misc irq handler loop (Gustavo) active phys bitmask and programming correction (Gustavo) v5: simplify pmdemand_state structure simplify methods to find active phys and max port clock Timeout in case of previou pmdemand task pending (Gustavo) v6: rebasing updates to max_ddiclk calculations (Gustavo) updates to active_phys count method (Gustavo) v7: use two separate loop to iterate throug old and new crtc states to calculate the active phys (Gustavo) v8: use uniform function names (Gustavo) v9: For phys change iterate through connectors (Imre) Look for change in phys for pmdemand update (Gustavo, Imre) Some more stlying changes (Imre) Update pmdemand state during HW readout/sanitize (Imre) v10: Fix CI checkpatch warnings v11: use correct pmdemand object pointer during hw readout, simplify the check for phys need update (Gustavo) v12: Handle possible non serialize cases (Imre) Initialise also pmdemand params HW readout (Imre) Update active phys mask during sanitize calls (Imre) Check TC/encoder changes to limit connector update (Imre) v13: Check display version before accessing pmdemand functions v14: Move is_serialized to intel_global_state.c simplify update params and other stlying issues (Imre) Bspec: 66451, 64636, 64602, 64603 Cc: Matt Atwood <matthew.s.atwood@intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: José Roberto de Souza <jose.souza@intel.com> Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: Mika Kahola <mika.kahola@intel.com> Signed-off-by: Vinod Govindapillai <vinod.govindapillai@intel.com> Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> #v4 Acked-by: Gustavo Sousa <gustavo.sousa@intel.com> #v11 Reviewed-by: Imre Deak <imre.deak@intel.com> [RK: Fixed minor typo in one of the comments. s/qclck_gc/qclk_gv/] Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230606201032.347449-1-vinod.govindapillai@intel.com
2023-06-06 23:10:32 +03:00
/* TODO move here (or even earlier?) on all platforms */
if (DISPLAY_VER(display) >= 9)
intel_wm_get_hw_state(display);
intel_bw_update_hw_state(display);
intel_cdclk_update_hw_state(display);
intel_pmdemand_init_pmdemand_params(display, pmdemand_state);
}
static void
get_encoder_power_domains(struct intel_display *display)
{
struct intel_encoder *encoder;
for_each_intel_encoder(display->drm, encoder) {
struct intel_crtc_state *crtc_state;
if (!encoder->get_power_domains)
continue;
/*
* MST-primary and inactive encoders don't have a crtc state
* and neither of these require any power domain references.
*/
if (!encoder->base.crtc)
continue;
crtc_state = to_intel_crtc_state(encoder->base.crtc->state);
encoder->get_power_domains(encoder, crtc_state);
}
}
static void intel_early_display_was(struct intel_display *display)
{
/*
* Display WA #1185 WaDisableDARBFClkGating:glk,icl,ehl,tgl
* Also known as Wa_14010480278.
*/
if (IS_DISPLAY_VER(display, 10, 12))
intel_de_rmw(display, GEN9_CLKGATE_DIS_0, 0, DARBF_GATING_DIS);
/*
* WaRsPkgCStateDisplayPMReq:hsw
* System hang if this isn't done before disabling all planes!
*/
if (display->platform.haswell)
intel_de_rmw(display, CHICKEN_PAR1_1, 0, FORCE_ARB_IDLE_PLANES);
if (display->platform.kabylake || display->platform.coffeelake ||
display->platform.cometlake) {
/* Display WA #1142:kbl,cfl,cml */
intel_de_rmw(display, CHICKEN_PAR1_1,
KBL_ARB_FILL_SPARE_22, KBL_ARB_FILL_SPARE_22);
intel_de_rmw(display, CHICKEN_MISC_2,
KBL_ARB_FILL_SPARE_13 | KBL_ARB_FILL_SPARE_14,
KBL_ARB_FILL_SPARE_14);
}
}
void intel_modeset_setup_hw_state(struct intel_display *display,
struct drm_modeset_acquire_ctx *ctx)
{
struct intel_encoder *encoder;
struct intel_crtc *crtc;
intel_wakeref_t wakeref;
wakeref = intel_display_power_get(display, POWER_DOMAIN_INIT);
intel_early_display_was(display);
intel_vga_disable(display);
intel_modeset_readout_hw_state(display);
/* HW state is read out, now we need to sanitize this mess. */
get_encoder_power_domains(display);
intel_pch_sanitize(display);
drm/i915/cmtg: Disable the CMTG The CMTG is a timing generator that runs in parallel with transcoders timing generators and can be used as a reference for synchronization. We have observed that we are inheriting from GOP a display configuration with the CMTG enabled. Because our driver doesn't currently implement any CMTG sequences, the CMTG ends up still enabled after our driver takes over. We need to make sure that the CMTG is not enabled if we are not going to use it. For that, let's add a partial implementation in our driver that only cares about disabling the CMTG if it was found enabled during initial hardware readout. In the future, we can also implement sequences for using the CMTG if that becomes a needed feature. For now, we only deal with cases when it is possible to disable the CMTG without requiring a modeset. For earlier display versions, we simply skip if we find the CMTG enabled and we can't disable it without a proper modeset. In the future, we need to properly handle that case. v2: - DG2 does not have the CMTG. Update HAS_CMTG() accordingly. - Update logic to force disabling of CMTG only for initial commit. v3: - Add missing changes for v2 that were staged but not committed. v4: - Avoid if/else duplication in intel_cmtg_dump_state() by using "n/a" for CMTG B enabled/disabled string for platforms without it. (Jani) - Prefer intel_cmtg_readout_hw_state() over intel_cmtg_readout_state(). (Jani) - Use display struct instead of i915 as first parameter for TRANS_DDI_FUNC_CTL2(). (Jani) - Fewer continuation lines in variable declaration/initialization for better readability. (Jani) - Coding style improvements. (Jani) - Use drm_dbg_kms() instead of drm_info() for logging the disabling of the CMTG. - Make struct intel_cmtg_state entirely private to intel_cmtg.c. v5: - Do the disable sequence as part of the sanitization step after hardware readout instead of initial modeset commit. (Jani) - Adapt to commit 15133582465f ("drm/i915/display: convert global state to struct intel_display") by using a display struct instead of i915 as argument for intel_atomic_global_obj_init(). v6: - Do not track CMTG state as a global state. (Ville) - Simplify the driver logic by only disabling the CMTG only on cases when a modeset is not required. (Ville) v7: - Remove the call to drm_WARN_ON() when checking intel_cmtg_disable_requires_modeset() and use a FIXME in the comment instead. - Remove the !HAS_CMTG() guard from intel_cmtg_get_config(), which is static and its caller is already protected by that same condition. - Also take the opportunity to put some Bspec references in the commit trailers section. v8: - Use HAS_TRANSCODER() instead of intel_crtc_for_pipe(). (Ville) - Ensure transcoder power well is enabled before reading TRANS_DDI_FUNC_CTL2. (Ville) Bspec: 68915, 49262 Cc: Jani Nikula <jani.nikula@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20250124173956.46534-1-gustavo.sousa@intel.com
2025-01-24 14:38:14 -03:00
intel_cmtg_sanitize(display);
/*
* intel_sanitize_plane_mapping() may need to do vblank
* waits, so we need vblank interrupts restored beforehand.
*/
for_each_intel_crtc(display->drm, crtc) {
struct intel_crtc_state *crtc_state =
to_intel_crtc_state(crtc->base.state);
intel_sanitize_fifo_underrun_reporting(crtc_state);
drm_crtc_vblank_reset(&crtc->base);
if (crtc_state->hw.active) {
intel_dmc_enable_pipe(crtc_state);
intel_crtc_vblank_on(crtc_state);
}
}
intel_fbc_sanitize(display);
intel_sanitize_plane_mapping(display);
for_each_intel_encoder(display->drm, encoder)
intel_sanitize_encoder(encoder);
/*
* Sanitizing CRTCs needs their connector atomic state to be
* up-to-date, so ensure that already here.
*/
intel_modeset_update_connector_atomic_state(display);
intel_sanitize_all_crtcs(display, ctx);
intel_dpll_sanitize_state(display);
/* TODO move earlier on all platforms */
if (DISPLAY_VER(display) < 9)
intel_wm_get_hw_state(display);
intel_wm_sanitize(display);
for_each_intel_crtc(display->drm, crtc) {
struct intel_crtc_state *crtc_state =
to_intel_crtc_state(crtc->base.state);
struct intel_power_domain_mask put_domains;
intel_modeset_get_crtc_power_domains(crtc_state, &put_domains);
if (drm_WARN_ON(display->drm, !bitmap_empty(put_domains.bits, POWER_DOMAIN_NUM)))
intel_modeset_put_crtc_power_domains(crtc, &put_domains);
}
intel_display_power_put(display, POWER_DOMAIN_INIT, wakeref);
intel_power_domains_sanitize_state(display);
}