[PATCH v12 00/16] Add HDMI 2.0 support to DW HDMI QP TX

[PATCH v12 00/16] Add HDMI 2.0 support to DW HDMI QP TX

[Cristian Ciocaltea]

Enable HDMI 2.0 display modes (e.g. 4K@60Hz) on the Synopsys DW HDMI QP
TX controller, as found in Rockchip RK3576 & RK3588 SoCs, by making use
of the scrambler infrastructure introduced recently in the DRM core and
HDMI display helpers.

The first group of patches wires up the SCDC scrambling feature itself
in the DW HDMI QP bridge driver, while improving i2c fault handling,
and providing enable/disable HPD PHY ops and a HPD notify helper.

The remaining patches cover the Rockchip platform driver and HPD
handling: bug fixes, minor cleanups, avoiding spurious HPD IRQ wakeups,
masking the RK3576 HPD IRQ in io_init, implementing the enable/disable
HPD PHY ops, switching HPD reports to dw_hdmi_qp_hpd_notify() to
restrict events to the affected connector, and finally dropping the
now-unused .setup_hpd() PHY op.

This has been tested on the following boards:

* Radxa ROCK 4D (RK3576)    - Cristian Ciocaltea
* Radxa ROCK 5B (RK3588)    - Diederik de Haas, Cristian Ciocaltea
* Orange Pi 5 (RK3588S)     - Robin Räber
* Orange Pi 5 Plus (RK3588) - Maud Spiering, Igor Paunovic
* NanoPC-T6 LTS (RK3588)    - Diederik de Haas

Regards,
Cristian

Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
---
Changes in v12:
- Rebased onto latest drm-misc-next and dropped the now-merged
  infrastructure patches
- Dropped all patches unrelated to dw-hdmi-qp: the vc4 and the other
  bridge changes have been already submitted as separate series, while
  the remaining kunit & final cleanup will follow
- Replaced patch "dw-hdmi-qp: Rate limit i2c read error messages" with
  "drm/bridge: dw-hdmi-qp: Demote i2c bus error messages to debug level"
- Updated "drm/bridge: dw-hdmi-qp: Provide dw_hdmi_qp_hpd_notify()
  helper", according to Maxime's review: hold a reference on the
  curr_conn connector while the bridge is enabled, take a temporary one
  in dw_hdmi_qp_hpd_notify(), and guard curr_conn with a mutex instead
  of READ_ONCE()/WRITE_ONCE()
- Collected Tested-by tags from Robin and Igor
- Link to v11: https://patch.msgid.link/20260901-dw-hdmi-qp-scramb-v11-0-bc12954a0688@collabora.com

Changes in v11:
- Collected R-b tags from Maxime and Heiko
- Restored connector status before -EDEADLK handling of
  drm_helper_probe_single_connector_modes() in "drm/probe-helper:
  Introduce .force_ctx() connector callback" (Sashiko)
- Improved "drm/connector: Add HDMI 2.0 scrambler infrastructure":
  * Constified drm_connector_hdmi_scrambler_supported()'s argument
  * Separated the checks for default tmds rates and the scrambler
    callbacks presence (Maxime)
  * Reworked drm_connector_hdmi_scrambler_supported() to rely
    exclusively on supported_hdmi_ver, and drop
    drmm_connector_hdmi_init()'s check for scrambler_{enable,disable}
    callbacks presence in non-HDMI2.0 cases
- Checked for connector->ddc before calling SCDC helpers from
  drm_connector_hdmi_disable_scrambling(), in "drm/display: hdmi: Add
  HDMI 2.0 scrambling management helpers" (Sashiko)
- Moved "drm/display: bridge_connector: Switch to .force_ctx() connector
  helper" patch after the one introducing
  drm_atomic_helper_connector_hdmi_force_ctx() to fix a build issue
- Added new patch "drm/bridge: Fix NULL deref in drm_bridge_add() for
  legacy bridges" (Sashiko)
- Split "drm/display: hdmi-state-helper: Pass acquire ctx to hotplug
  helpers" into "drm/display: hdmi-state-helper: Add
  drm_atomic_helper_connector_hdmi_force_ctx()" (Maxime)
- Improved drm_atomic_helper_connector_hdmi_update()'s error handling in
  "drm/display: hdmi-state-helper: Sync SCDC state on hotplug" (Maxime)
- Reordered cancel_delayed_work_sync() and dw_hdmi_qp_suspend()
  in dw_hdmi_qp_rockchip_suspend() and prevented I2C timeouts and bus
  hangs during system suspend, part of "drm/rockchip: dw_hdmi_qp: Cancel
  pending HPD work on suspend" (Sashiko)
- Canceled an already queued hdmi->hpd_work after disabling the HPD IRQ
  and thus preventing any spurious HPD events in "drm/rockchip:
  dw_hdmi_qp: Control the HPD IRQ line via the bridge HPD ops" (Sashiko)
- Ignored failures when the scrambler hooks try to acquire the drm
  device, in "drm/vc4: hdmi: Convert to common HDMI 2.0 scrambling
  infrastructure" (Maxime)
- Moved the comment describing the vendor/product strings to the correct
  position in "drm/sun4i: hdmi: Convert to drmm_connector_hdmi_init()"
  (Sashiko)
- Updated "drm/connector: Drop redundant hdmi vendor/product fields"
  description to indicate the connector registration tests used to
  verify the zero-padded product/vendor fields will be covered by the
  SPD InfoFrame tests
- Provided a parameterized test for the missing scrambler callbacks and
  drop the now redundant tests in "drm/tests: connector: Add HDMI
  source-side scrambler coverage"
- Added new patches to consolidate SPD InfoFrame handling of
  vendor/product:
  * video/hdmi: Define SPD InfoFrame field lengths and use
    strtomem_pad()
  * drm/connector: Use the SPD InfoFrame field length defines
  * drm/tests: hdmi: Add SPD InfoFrame vendor/product coverage
- Rebased series onto latest drm-misc-next; since the tree now contains
  a workaround for the broken EL2 virtual timer on bcm2712 SoC, dropped
  the cover letter note about RPi5 hang during boot
- Link to v10:
  https://patch.msgid.link/20260731-dw-hdmi-qp-scramb-v10-0-294364b2cf15@collabora.com

Changes in v10:
- Collected T-b from Diederik on relevant patches
- Collected R-b from Dmitry on patch renaming drmm_connector_hdmi_init()
- Added new patches to address Sashiko findings:
  * drm/connector: hdmi: Handle reset() state allocation failure
  * drm/bridge: Fix unlocked list_del in drm_bridge_add()
  * drm/bridge: Fix unlocked list access in drm_bridge_attach()
  * drm/rockchip: dw_hdmi_qp: Cancel pending HPD work on suspend
  * drm/probe-helper: Introduce .force_ctx() connector callback
  * drm/display: bridge_connector: Switch to .force_ctx() connector
    helper
  * drm/vc4: hdmi: Switch to .force_ctx() connector helper
  * drm/display: hdmi-state-helper: Drop
    drm_atomic_helper_connector_hdmi_force()
- Updated "drm/connector: Add drmm_connector_hdmi_init() with new
  signature" to check if the reset callback failed to allocate memory
- Updated "drm/display: hdmi-state-helper: Pass acquire ctx to hotplug
  helpers" to introduce drm_atomic_helper_connector_hdmi_force_ctx()
  implementation and drop the conflicting local ctx from
  drm_atomic_helper_connector_hdmi_force()
- Updated "drm/display: hdmi: Add HDMI 2.0 scrambling management
  helpers" to bail early in the work handler on
  drm_connector_is_unregistered()
- Updated "drm/bridge: dw-hdmi-qp: Provide dw_hdmi_qp_hpd_notify()
  helper" to use READ_ONCE/WRITE_ONCE on hdmi->curr_conn
- Reworked "drm/rockchip: dw_hdmi_qp: Fix NULL deref in PM ops on
  incomplete bind" into "drm/rockchip: dw_hdmi_qp: Fix invalid drvdata
  access in PM ops"
- Fixed "drm/rockchip: rk3066_hdmi: Advertise HDMI 1.4 capabilities" to
  set the expected HDMI_VERSION_1_4 instead of HDMI_VERSION_1_3
- Fixed all KUnit tests that created a dangling pointer by storing the
  address of a stack-allocated drm_connector_hdmi_funcs structure into a
  persistent device-managed object
- Fixed flawed logic in "drm/tests: connector: Add HDMI
  max_tmds_char_rate init coverage" causing a couple of rate limit
  rejection tests to pass for the wrong reason
- Rebased series onto latest drm-misc-next
- Link to v9: https://patch.msgid.link/20260723-dw-hdmi-qp-scramb-v9-0-4fb12ea22ac9@collabora.com

Changes in v9:
- Collected A-b, R-b, T-b tags from Dmitry, Maxime, Maud
- Dropped misleading "HDMI_VERSION_UNKNOWN means legacy/default
  behaviour" from drm_connector_hdmi_caps.supported_hdmi_ver
  documentation
- Warned in drm_bridge_add() if an HDMI bridge leaves supported_hdmi_ver
  unset
- Added max_tmds_char_rate to drm_bridge struct and wire it through
  drm_bridge_connector_init() along with supported_hdmi_ver
- Got rid of connector's scrambler_supported flag and, instead, relied
  on checking the presence of scrambler callbacks via a new
  drm_connector_hdmi_scrambler_supported() helper (Dmitry)
- Dropped HDMI connector scdc_cb callback and introduced
  scdc_work_initialized flag to perform lazy initialization and guard
  the teardown paths against touching an uninitialized work item
- Unconditionally checked connector->hdmi.max_tmds_char_rate in
  "drm/display: hdmi-state-helper: Add fallback TMDS rate validation"
  and performed s/fallback/source/ in commit subject & description
  (Maxime)
- Added a TODO comment for checking
  drm_display_info.hdmi.scdc.scrambling.low_rates in "drm/display:
  hdmi-state-helper: Set HDMI scrambling requirement" patch (Maxime)
- Moved the prototype changes from "drm/display: hdmi-state-helper: Sync
  SCDC state on hotplug" into a dedicated "drm/display:
  hdmi-state-helper: Pass acquire ctx to hotplug helpers" patch; also
  switched ctx and status args ordering, for consistency with
  detect_ctx() signature (Maxime)
- Ensured HDMI connector's SCDC delayed work is explicitly canceled
  during connector cleanup (Sashiko)
- Fixed a potential NULL pointer dereference in
  drm_connector_hdmi_enable_scrambling() and
  drm_connector_hdmi_sync_scdc() helpers caused by forcing an EDID
  advertising SCDC on a connector that physically lacks an I2C DDC bus
  (Sashiko)
- dw-hdmi-qp
  * Fixed a potential NULL pointer dereference in
    dw_hdmi_qp_hpd_notify() helper by using a local variable to cache
    the connector pointer (Sashiko)
  * Ensured hdmi->phy.ops->disable() is not skipped in
    dw_hdmi_qp_bridge_atomic_disable() by accessing the connector via
    drm_atomic_get_old_connector_for_encoder() instead of
    hdmi->curr_conn (Sashiko)
  * Mitigated a potential use-after-free reported by Sashiko caused when
    HPD interrupt fires after the DRM device is torn down; reworked
    "Defer HPD IRQ enable until after connector setup" into:
    * Factor out HPD interrupt mask/unmask helpers
    * Control the HPD IRQ line via the bridge HPD ops
  * Added new patch addressing another Sashiko finding: drm/rockchip:
    dw_hdmi_qp: Avoid spurious HPD IRQ thread wakeups
  * Added new patch: drm/bridge: dw-hdmi-qp: Remove unused workqueue
    include and define
  * Dropped .hdmi_tmds_char_rate_valid() usage from the
    bridge/synopsys/dw-hdmi-qp.c scrambling support patch, while
    advertising HDMI 1.4 capability for no-hpd configurations
- VC4
  * Passed ctx instead of NULL when calling
    drm_atomic_helper_connector_hdmi_hotplug() from
    vc4_hdmi_handle_hotplug() in the helper preparatory patch
  * Aligned with the expected detect_ctx() semantics in a dedicated
    "drm/vc4: hdmi: Propagate -EDEADLK to the top level" rather than
    bundled in the conversion patch
  * Optimized diff around drm_dev_enter() in the conversion patch
    (Maxime)
  * Simplified vc4_hdmi_connector_clock_valid() logic by deferring max
    clock validation to HDMI helpers (new patch)
- Migrated to a funcs-based HDMI connector init API - new patches:
  * drm/display: hdmi: Rename drmm_connector_hdmi_init() to *_ini2()
  * drm/display: bridge_connector: Convert to drmm_connector_hdmi_init()
  * drm/vc4: hdmi: Convert to drmm_connector_hdmi_init()
  * drm/vc4: hdmi: Convert to common HDMI 2.0 scrambling infrastructure
  * drm/sun4i: hdmi: Convert to drmm_connector_hdmi_init()
  * drm/tests: connector: Convert to drmm_connector_hdmi_init()
  * drm/tests: hdmi_state_helper: Convert to drmm_connector_hdmi_init()
  * drm/bridge: adv7511: Advertise HDMI 1.2 capabilities
  * drm/bridge: inno-hdmi: Advertise HDMI 1.2 capabilities
  * drm/bridge: ite-it6263: Drop redundant .mode_valid hook
  * drm/bridge: ite-it6263: Advertise HDMI 1.3 capabilities
  * drm/bridge: ite-it66121: Advertise HDMI 1.2 capabilities
  * drm/bridge: lontium-lt9611: Advertise HDMI 1.4 capabilities
  * drm/rockchip: rk3066_hdmi: Advertise HDMI 1.4 capabilities
  * drm/sun4i: hdmi: Convert to drmm_connector_hdmi_init()
  * drm/connector: Remove drmm_connector_hdmi_ini2()
  * drm/connector: Drop redundant hdmi vendor/product fields
  * drm/connector: Drop redundant hdmi supported_formats field
  * drm/connector: Drop redundant max_bpc field
- KUnit
  * Adapted hdmi_state_helper scrambler tests and reworked a couple of
    max_tmds_char_rate tests according to the latest DRM core changes
  * Fixed a regression in the HDMI caps-based init coverage
  * Fix a couple of bisectability issues, also implied partial commit
    reorder
  * Set supported HDMI version for the test bridges (new patch)
- Rebased series onto latest drm-misc-next and removed "drm/rockchip:
  dw_hdmi_qp: Drop unnecessary #include" patch, since a similar change
  as been already applied
- Link to v8: https://patch.msgid.link/20260702-dw-hdmi-qp-scramb-v8-0-d79890d00b6a@collabora.com

Changes in v8:
- Added an enum hdmi_version under video/hdmi to describe the HDMI
  specification version a source/sink supports (new patch 1)
- Introduced HDMI connector capabilities API and reworked the
  infrastructure patches on top of it: drmm_connector_hdmi_init() now
  takes a drm_connector_hdmi_caps struct that carries supported_formats,
  max_bpc, supported_hdmi_ver and max_tmds_char_rate (new patches 2-3)
- Moved all the new SCDC helpers covering both source and sink to
  display/drm_hdmi_helper.c
- Renamed drm_scdc_{start,stop}_scrambling() to
  drm_connector_hdmi_{enable,disable}_scrambling() and added a
  connector-state argument so the scrambler_needed flag is checked
  centrally instead of in every driver
- Renamed drm_scdc_sync_status() to drm_connector_hdmi_sync_scdc(),
  aligned its lifecycle with vc4_hdmi_reset_link(), and removed the
  internal drm_scdc_reset_crtc() helper; it now also gates on CRTC state
  and an in-flight commit (Maxime)
- Changed the prototype of drm_atomic_helper_connector_hdmi_hotplug() to
  take an acquire context and an int return, instead of adding a new
  _ctx variant (Maxime)
- Added drm_hdmi_mode_needs_scrambling() to centralise the above-340 MHz
  decision (new patch 7)
- Made SCDC source version negotiation optional and moved it to a
  dedicated patch "drm/display: hdmi: Advertise SCDC source version when
  scrambling" (new patch 8), backed by a new
  drm_scdc_set_source_version() helper in "drm/display: scdc-helper: Add
  helper to set SCDC version information" (new patch 6)
- Added a fallback TMDS rate validation patch using the connector-level
  max_tmds_char_rate when the driver provides no .tmds_char_rate_valid()
  hook (new patch 9)
- Split scrambler_needed flag handling into its own hdmi-state-helper
  dedicate change (new patch 11)
- VC4:
  * Replaced vc4_hdmi_mode_needs_scrambling() with
    drm_hdmi_mode_needs_scrambling() (new patch 30)
  * Restored drm_dev_{enter,exit}() pairing around the converted
    scrambling paths
  * Removed now unused output_{bpc,color_format} fields from vc4_hdmi as
    part of the SCDC conversion
- KUnit:
  * Covered the caps-based connector init path: NULL caps, inferred
    max_tmds_char_rate per supported_hdmi_ver, and override validation
    (new patch 32)
  * Switched drm_hdmi_state_helper_test to the caps-based init helper
    and exposed it through
    drm_kunit_helper_connector_hdmi_init_with_caps_edid_funcs() (new
    patch 34)
  * Added max_tmds_char_rate fallback coverage in
    drm_hdmi_state_helper_test (new patch 35)
- Rebased onto latest drm-misc-next and dropped the already applied
  "drm/fb-helper: Remove unused local variable in hotplug_event()" patch
- Link to v7: https://patch.msgid.link/20260602-dw-hdmi-qp-scramb-v7-0-445eb54ee1ed@collabora.com

Changes in v7:
- Split the scrambler_needed flag computation from the SCDC scrambling
  helpers into a standalone hdmi_state_helper patch for clarity (new
  patch 6)
- Added vc4 HDMI conversion to common TMDS char rate constants and SCDC
  scrambling helpers as a proof of reuse (new patches 24-25)
- Added KUnit tests for connector scrambler_supported/callback
  consistency validation (new patch 26)
- Added 4K@60Hz 600MHz TMDS test EDID for high-rate scrambling coverage
  (new patch 27)
- Added KUnit tests for hdmi_state_helper scrambler_needed flag:
  low-rate, high-rate, and source-no-support cases (new patch 28)
- Fixed EDID conformity issues (edid-decode v1.33.0 failures) in the
  existing 1080p+4K YUV420 200MHz and 4K RGB/YUV 340MHz test EDIDs
  without changing the capabilities used by existing tests (new patches
  29-30)
- Rebased onto latest drm-misc-next
- Link to v6: https://patch.msgid.link/20260520-dw-hdmi-qp-scramb-v6-0-24b74603b782@collabora.com

Changes in v6:
- Collected R-b & A-b tags from Dmitry and Heiko
- Restructured the series from 10 into 22 patches, splitting the SCDC
  scrambling helpers, connector infrastructure, bridge operations,
  bridge_connector plumbing, dw-hdmi-qp implementation, and Rockchip
  platform changes into distinct commits
- Added drm_scdc_dbg() macro to simplify debug messages (new patch)
- Added drm_scdc_start/stop/sync_status() helpers for full scrambling
  lifecycle management (reworked from v5 patch 5)
- Added drm_atomic_helper_connector_hdmi_hotplug_ctx() to propagate
  modeset acquire context for SCDC sync on hotplug (new patch)
- Added connector scrambler callbacks and SCDC work/flag infrastructure
  as a separate patch (split from v5 patch 5)
- Added DRM_BRIDGE_OP_HDMI_SCRAMBLER bridge operation with
  hdmi_scrambler_enable/disable callbacks (new patch)
- Implemented bridge_connector scrambler interface wiring (new patch)
- Added .enable_hpd()/.disable_hpd() PHY ops for dw-hdmi-qp bridge and
  Rockchip platform drivers, replacing the obsolete .setup_hpd() op
- Added dw_hdmi_qp_hpd_notify() helper for targeted connector-only HPD
  notification (split from v5 patch 10)
- Dropped drm_fb_helper_hotplug_event() unused variable (new cleanup)
- Dropped unused drm_simple_kms_helper.h include (new cleanup)
- Masked RK3576 HPD IRQ in io_init() for consistency with RK3588
- Rebased onto latest drm-misc-next
- Link to v5: https://patch.msgid.link/20260426-dw-hdmi-qp-scramb-v5-0-d778e70c317b@collabora.com

Changes in v5:
- Added new patches: 1/10, 3/10, 6/10, 7/10, 8/10
- Removed redundant no-op error check in drm_bridge_helper_reset_crtc()
  (patch 1)
- Removed the EDEADLK retry loop from the bridge .detect_ctx() callback,
  as that's already handled in the drm_bridge_detect_ctx() helper or by
  the caller when ctx is provided (patch 2)
- Refactored drm_bridge_detect() to delegate to drm_bridge_detect_ctx()
  and added a WARN_ON for unexpected negative return values (patch 2)
- Split the bridge-connector .detect_ctx() switch into a preparatory
  patch to use cached connector status in .get_modes() (patch 3)
- Improved error handling in SCDC scrambling setup: roll back high TMDS
  clock ratio on scrambling failure, reset scramb_enabled flag on
  set_scramb failure, and add SCDC version read/write error checks
  (patch 5)
- Annotated scramb_enabled with READ_ONCE/WRITE_ONCE for cross-context
  access between modeset paths and the scrambling work item (patch 5)
- Renamed SCDC_MIN_SOURCE_VERSION to SCDC_MAX_SOURCE_VERSION (patch 5)
- Rate limited i2c error messages (patch 6)
- Added missing newlines in dev_err_probe() messages (patch 7)
- Replaced indirect device pointer accesses with local dev variable in
  bind() (patch 8)
- Split the HPD connector restriction (formerly patch 4/4): register HPD
  IRQ after connector setup first (patch 9), then restrict HPD event to
  the affected connector (patch 10); also collected R-b from Heiko
- Rebased onto latest drm-misc-next
- Link to v4: https://lore.kernel.org/r/20260303-dw-hdmi-qp-scramb-v4-0-317d3b8bd219@collabora.com

Changes in v4:
- Fixed conflicts while rebasing onto latest drm-misc-next
- Link to v3: https://lore.kernel.org/r/20260119-dw-hdmi-qp-scramb-v3-0-bd8611730fc1@collabora.com

Changes in v3:
- Used drm_bridge_helper_reset_crtc() helper to reset the display
  pipeline and got rid of some boilerplate code (Maxime)
- Rebased onto latest drm-misc-next
- Link to v2: https://lore.kernel.org/r/20260113-dw-hdmi-qp-scramb-v2-0-ae7b2c58d24d@collabora.com

Changes in v2:
- Collected Tested-by tags from Diederik and Maud
- Rebased onto latest drm-misc-next
- Ensured the recently introduced 'no-hpd' support for dealing with
  unconnected/repurposed/broken HPD pin is limited to HDMI 1.4 rates
- Link to v1: https://lore.kernel.org/r/20251203-dw-hdmi-qp-scramb-v1-0-836fe7401a69@collabora.com

---
Cristian Ciocaltea (16):
      drm/bridge: dw-hdmi-qp: Demote i2c bus error messages to debug level
      drm/bridge: dw-hdmi-qp: Provide .{enable,disable}_hpd() PHY ops
      drm/bridge: dw-hdmi-qp: Remove unused workqueue include and define
      drm/bridge: dw-hdmi-qp: Add HDMI 2.0 scrambling support
      drm/bridge: dw-hdmi-qp: Provide dw_hdmi_qp_hpd_notify() helper
      drm/rockchip: dw_hdmi_qp: Fix invalid drvdata access in PM ops
      drm/rockchip: dw_hdmi_qp: Cancel pending HPD work on suspend
      drm/rockchip: dw_hdmi_qp: Add missing newlines in dev_err_probe() messages
      drm/rockchip: dw_hdmi_qp: Use local dev variable consistently in bind()
      drm/rockchip: dw_hdmi_qp: Avoid spurious HPD IRQ thread wakeups
      drm/rockchip: dw_hdmi_qp: Mask RK3576 HPD IRQ in io_init
      drm/rockchip: dw_hdmi_qp: Implement .{enable,disable}_hpd() PHY ops
      drm/rockchip: dw_hdmi_qp: Factor out HPD interrupt (un)mask helpers
      drm/rockchip: dw_hdmi_qp: Control the HPD IRQ line via the bridge HPD ops
      drm/rockchip: dw_hdmi_qp: Use dw_hdmi_qp_hpd_notify() for HPD reports
      drm/bridge: dw-hdmi-qp: Drop unused .setup_hpd() phy op

 drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c   | 159 ++++++++++++++-----
 drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c | 205 +++++++++++++++----------
 include/drm/bridge/dw_hdmi_qp.h                |   4 +-
 3 files changed, 248 insertions(+), 120 deletions(-)
---
base-commit: 37b346eab23aead52d17a8fa687056bfc651a416
change-id: 20251203-dw-hdmi-qp-scramb-cdbd8b57ccf9

[PATCH v12 01/16] drm/bridge: dw-hdmi-qp: Demote i2c bus error messages to debug level

[Cristian Ciocaltea]

A NACK or timeout on the DDC bus is a normal condition when the sink is
absent or unpowered.  Some displays keep HPD asserted and the EDID
EEPROM readable while in standby, but leave the SCDC slave unpowered.
This results in i2c-related error messages on every boot during
scrambling setup, even though the condition is handled gracefully and
the SCDC configuration is restored on the next hotplug event.

Log the i2c bus level failures with dev_dbg_ratelimited() instead,
matching the silent behaviour of the non-QP dw-hdmi i2c bus algorithm.
The DDC users already report failures at the appropriate level, e.g. the
SCDC helpers via drm_scdc_dbg().  This also makes the no-hpd special
casing of the read path messages redundant, so drop it.

Reported-by: Diederik de Haas <diederik@cknow-tech.com>
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
---
 drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c | 16 ++++------------
 1 file changed, 4 insertions(+), 12 deletions(-)

diff --git a/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c b/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c
index c9568aea98dc..9abcea92b720 100644
--- a/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c
+++ b/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c
@@ -350,22 +350,14 @@ static int dw_hdmi_qp_i2c_read(struct dw_hdmi_qp *hdmi,
 
 		stat = wait_for_completion_timeout(&i2c->cmp, HZ / 10);
 		if (!stat) {
-			if (hdmi->no_hpd)
-				dev_dbg_ratelimited(hdmi->dev,
-						    "i2c read timed out\n");
-			else
-				dev_err(hdmi->dev, "i2c read timed out\n");
+			dev_dbg_ratelimited(hdmi->dev, "i2c read timed out\n");
 			dw_hdmi_qp_write(hdmi, 0x01, I2CM_CONTROL0);
 			return -EAGAIN;
 		}
 
 		/* Check for error condition on the bus */
 		if (i2c->stat & I2CM_NACK_RCVD_IRQ) {
-			if (hdmi->no_hpd)
-				dev_dbg_ratelimited(hdmi->dev,
-						    "i2c read error\n");
-			else
-				dev_err(hdmi->dev, "i2c read error\n");
+			dev_dbg_ratelimited(hdmi->dev, "i2c read error\n");
 			dw_hdmi_qp_write(hdmi, 0x01, I2CM_CONTROL0);
 			return -EIO;
 		}
@@ -404,14 +396,14 @@ static int dw_hdmi_qp_i2c_write(struct dw_hdmi_qp *hdmi,
 
 		stat = wait_for_completion_timeout(&i2c->cmp, HZ / 10);
 		if (!stat) {
-			dev_err(hdmi->dev, "i2c write time out!\n");
+			dev_dbg_ratelimited(hdmi->dev, "i2c write time out!\n");
 			dw_hdmi_qp_write(hdmi, 0x01, I2CM_CONTROL0);
 			return -EAGAIN;
 		}
 
 		/* Check for error condition on the bus */
 		if (i2c->stat & I2CM_NACK_RCVD_IRQ) {
-			dev_err(hdmi->dev, "i2c write nack!\n");
+			dev_dbg_ratelimited(hdmi->dev, "i2c write nack!\n");
 			dw_hdmi_qp_write(hdmi, 0x01, I2CM_CONTROL0);
 			return -EIO;
 		}

-- 
2.55.0

[PATCH v12 02/16] drm/bridge: dw-hdmi-qp: Provide .{enable,disable}_hpd() PHY ops

[Cristian Ciocaltea]

Implement the .hpd_enable() and .hpd_disable() bridge callbacks and
extend dw_hdmi_qp_phy_ops with corresponding hooks.

This enables the DRM core to control when HPD interrupts are armed,
which is needed to prevent premature interrupt delivery before the
connector is fully initialized, and to properly quiesce HPD during
suspend.

Tested-by: Maud Spierings <maud_spierings@hotmail.com>
Tested-by: Diederik de Haas <diederik@cknow-tech.com>  # NanoPC-T6 LTS, Rock 5B
Tested-by: Robin Räber <robin.raebi@gmail.com>
Tested-by: Igor Paunovic <royalnet026@gmail.com> # RK3588 Orange Pi 5 Plus, dual HDMI
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
---
 drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c | 18 ++++++++++++++++++
 include/drm/bridge/dw_hdmi_qp.h              |  2 ++
 2 files changed, 20 insertions(+)

diff --git a/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c b/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c
index 9abcea92b720..bed743ce3909 100644
--- a/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c
+++ b/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c
@@ -974,6 +974,22 @@ static int dw_hdmi_qp_cec_transmit(struct drm_bridge *bridge, u8 attempts,
 #define dw_hdmi_qp_cec_transmit NULL
 #endif /* CONFIG_DRM_DW_HDMI_QP_CEC */
 
+static void dw_hdmi_qp_bridge_hpd_enable(struct drm_bridge *bridge)
+{
+	struct dw_hdmi_qp *hdmi = bridge->driver_private;
+
+	if (hdmi->phy.ops->enable_hpd)
+		hdmi->phy.ops->enable_hpd(hdmi, hdmi->phy.data);
+}
+
+static void dw_hdmi_qp_bridge_hpd_disable(struct drm_bridge *bridge)
+{
+	struct dw_hdmi_qp *hdmi = bridge->driver_private;
+
+	if (hdmi->phy.ops->disable_hpd)
+		hdmi->phy.ops->disable_hpd(hdmi, hdmi->phy.data);
+}
+
 static const struct drm_bridge_funcs dw_hdmi_qp_bridge_funcs = {
 	.atomic_get_output_bus_fmts = drm_atomic_helper_bridge_get_hdmi_output_bus_fmts,
 	.atomic_duplicate_state = drm_atomic_helper_bridge_duplicate_state,
@@ -982,6 +998,8 @@ static const struct drm_bridge_funcs dw_hdmi_qp_bridge_funcs = {
 	.atomic_enable = dw_hdmi_qp_bridge_atomic_enable,
 	.atomic_disable = dw_hdmi_qp_bridge_atomic_disable,
 	.detect = dw_hdmi_qp_bridge_detect,
+	.hpd_enable = dw_hdmi_qp_bridge_hpd_enable,
+	.hpd_disable = dw_hdmi_qp_bridge_hpd_disable,
 	.edid_read = dw_hdmi_qp_bridge_edid_read,
 	.hdmi_tmds_char_rate_valid = dw_hdmi_qp_bridge_tmds_char_rate_valid,
 	.hdmi_clear_avi_infoframe = dw_hdmi_qp_bridge_clear_avi_infoframe,
diff --git a/include/drm/bridge/dw_hdmi_qp.h b/include/drm/bridge/dw_hdmi_qp.h
index 6ea9c561cfef..b80fceffc315 100644
--- a/include/drm/bridge/dw_hdmi_qp.h
+++ b/include/drm/bridge/dw_hdmi_qp.h
@@ -17,6 +17,8 @@ struct dw_hdmi_qp_phy_ops {
 	void (*disable)(struct dw_hdmi_qp *hdmi, void *data);
 	enum drm_connector_status (*read_hpd)(struct dw_hdmi_qp *hdmi, void *data);
 	void (*setup_hpd)(struct dw_hdmi_qp *hdmi, void *data);
+	void (*enable_hpd)(struct dw_hdmi_qp *hdmi, void *data);
+	void (*disable_hpd)(struct dw_hdmi_qp *hdmi, void *data);
 };
 
 struct dw_hdmi_qp_plat_data {

-- 
2.55.0

[PATCH v12 03/16] drm/bridge: dw-hdmi-qp: Remove unused workqueue include and define

[Cristian Ciocaltea]

linux/workqueue.h is no longer needed, and SCRAMB_POLL_DELAY_MS is
unused.

Remove them.

Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
---
 drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c b/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c
index bed743ce3909..4b55bed42ab7 100644
--- a/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c
+++ b/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c
@@ -15,7 +15,6 @@
 #include <linux/module.h>
 #include <linux/mutex.h>
 #include <linux/of.h>
-#include <linux/workqueue.h>
 
 #include <drm/bridge/dw_hdmi_qp.h>
 #include <drm/display/drm_hdmi_helper.h>
@@ -38,8 +37,6 @@
 #define DDC_CI_ADDR		0x37
 #define DDC_SEGMENT_ADDR	0x30
 
-#define SCRAMB_POLL_DELAY_MS	3000
-
 struct dw_hdmi_qp_i2c {
 	struct i2c_adapter	adap;
 

-- 
2.55.0

[PATCH v12 04/16] drm/bridge: dw-hdmi-qp: Add HDMI 2.0 scrambling support

[Cristian Ciocaltea]

Add support for SCDC scrambling and the high TMDS clock ratio required
by HDMI 2.0 modes with TMDS character rates above 340 MHz, such as
4K@60Hz.

Hook the bridge into the DRM HDMI connector scrambling helpers so the
sink-side SCDC state and the controller scrambler are updated together
as part of the atomic enable and disable paths.  This allows the driver
to advertise HDMI 2.0 support when HPD is available.

Keep no-HPD configurations limited to HDMI 1.4, since SCDC access
requires a connected HDMI sink.

Modes above the HDMI 2.0 600 MHz TMDS character-rate limit are now
implicitly rejected by the DRM HDMI connector/bridge infrastructure,
making the dw_hdmi_qp_bridge_tmds_char_rate_valid() callback redundant.
Drop it.

Tested-by: Maud Spierings <maud_spierings@hotmail.com>
Tested-by: Diederik de Haas <diederik@cknow-tech.com>  # NanoPC-T6 LTS, Rock 5B
Acked-by: Heiko Stuebner <heiko@sntech.de>
Tested-by: Robin Räber <robin.raebi@gmail.com>
Tested-by: Igor Paunovic <royalnet026@gmail.com> # RK3588 Orange Pi 5 Plus, dual HDMI
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
---
 drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c | 60 ++++++++++++++++++----------
 1 file changed, 39 insertions(+), 21 deletions(-)

diff --git a/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c b/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c
index 4b55bed42ab7..14723ee5cb05 100644
--- a/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c
+++ b/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c
@@ -2,6 +2,7 @@
 /*
  * Copyright (c) 2021-2022 Rockchip Electronics Co., Ltd.
  * Copyright (c) 2024 Collabora Ltd.
+ * Copyright (c) 2025 Amazon.com, Inc. or its affiliates.
  *
  * Author: Algea Cao <algea.cao@rock-chips.com>
  * Author: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
@@ -536,6 +537,7 @@ static void dw_hdmi_qp_bridge_atomic_enable(struct drm_bridge *bridge,
 	struct drm_connector_state *conn_state;
 	struct drm_connector *connector;
 	unsigned int op_mode;
+	int ret;
 
 	connector = drm_atomic_get_new_connector_for_encoder(state, bridge->encoder);
 	if (WARN_ON(!connector))
@@ -546,14 +548,20 @@ static void dw_hdmi_qp_bridge_atomic_enable(struct drm_bridge *bridge,
 		return;
 
 	if (connector->display_info.is_hdmi) {
-		dev_dbg(hdmi->dev, "%s mode=HDMI %s rate=%llu bpc=%u\n", __func__,
-			drm_hdmi_connector_get_output_format_name(conn_state->hdmi.output_format),
-			conn_state->hdmi.tmds_char_rate, conn_state->hdmi.output_bpc);
 		op_mode = 0;
 		hdmi->tmds_char_rate = conn_state->hdmi.tmds_char_rate;
+
+		ret = drm_connector_hdmi_enable_scrambling(connector, conn_state);
+		if (ret)
+			dev_warn(hdmi->dev, "Failed to enable scrambling: %d\n", ret);
+
+		dev_dbg(hdmi->dev, "%s mode=HDMI %s rate=%llu bpc=%u scramb=%d\n", __func__,
+			drm_hdmi_connector_get_output_format_name(conn_state->hdmi.output_format),
+			conn_state->hdmi.tmds_char_rate, conn_state->hdmi.output_bpc,
+			connector->hdmi.scrambler_enabled);
 	} else {
-		dev_dbg(hdmi->dev, "%s mode=DVI\n", __func__);
 		op_mode = OPMODE_DVI;
+		dev_dbg(hdmi->dev, "%s mode=DVI\n", __func__);
 	}
 
 	hdmi->phy.ops->init(hdmi, hdmi->phy.data);
@@ -568,9 +576,13 @@ static void dw_hdmi_qp_bridge_atomic_disable(struct drm_bridge *bridge,
 					     struct drm_atomic_commit *state)
 {
 	struct dw_hdmi_qp *hdmi = bridge->driver_private;
+	struct drm_connector *connector;
 
 	hdmi->tmds_char_rate = 0;
 
+	connector = drm_atomic_get_old_connector_for_encoder(state, bridge->encoder);
+	drm_connector_hdmi_disable_scrambling(connector);
+
 	hdmi->phy.ops->disable(hdmi, hdmi->phy.data);
 }
 
@@ -605,24 +617,24 @@ dw_hdmi_qp_bridge_edid_read(struct drm_bridge *bridge,
 	return drm_edid;
 }
 
-static enum drm_mode_status
-dw_hdmi_qp_bridge_tmds_char_rate_valid(const struct drm_bridge *bridge,
-				       const struct drm_display_mode *mode,
-				       unsigned long long rate)
+static int dw_hdmi_qp_bridge_scrambler_enable(struct drm_bridge *bridge)
 {
 	struct dw_hdmi_qp *hdmi = bridge->driver_private;
 
-	/*
-	 * TODO: when hdmi->no_hpd is 1 we must not support modes that
-	 * require scrambling, including every mode with a clock above
-	 * HDMI_1_3_TMDS_CHAR_RATE_MAX_HZ.
-	 */
-	if (rate > HDMI_1_3_TMDS_CHAR_RATE_MAX_HZ) {
-		dev_dbg(hdmi->dev, "Unsupported TMDS char rate: %lld\n", rate);
-		return MODE_CLOCK_HIGH;
-	}
+	dw_hdmi_qp_write(hdmi, 1, SCRAMB_CONFIG0);
+	dev_dbg(hdmi->dev, "scrambler enabled\n");
 
-	return MODE_OK;
+	return 0;
+}
+
+static int dw_hdmi_qp_bridge_scrambler_disable(struct drm_bridge *bridge)
+{
+	struct dw_hdmi_qp *hdmi = bridge->driver_private;
+
+	dw_hdmi_qp_write(hdmi, 0, SCRAMB_CONFIG0);
+	dev_dbg(hdmi->dev, "scrambler disabled\n");
+
+	return 0;
 }
 
 static int dw_hdmi_qp_bridge_clear_avi_infoframe(struct drm_bridge *bridge)
@@ -998,7 +1010,8 @@ static const struct drm_bridge_funcs dw_hdmi_qp_bridge_funcs = {
 	.hpd_enable = dw_hdmi_qp_bridge_hpd_enable,
 	.hpd_disable = dw_hdmi_qp_bridge_hpd_disable,
 	.edid_read = dw_hdmi_qp_bridge_edid_read,
-	.hdmi_tmds_char_rate_valid = dw_hdmi_qp_bridge_tmds_char_rate_valid,
+	.hdmi_scrambler_enable = dw_hdmi_qp_bridge_scrambler_enable,
+	.hdmi_scrambler_disable = dw_hdmi_qp_bridge_scrambler_disable,
 	.hdmi_clear_avi_infoframe = dw_hdmi_qp_bridge_clear_avi_infoframe,
 	.hdmi_write_avi_infoframe = dw_hdmi_qp_bridge_write_avi_infoframe,
 	.hdmi_clear_hdmi_infoframe = dw_hdmi_qp_bridge_clear_hdmi_infoframe,
@@ -1126,13 +1139,18 @@ struct dw_hdmi_qp *dw_hdmi_qp_bind(struct platform_device *pdev,
 			   DRM_BRIDGE_OP_HDMI_AUDIO |
 			   DRM_BRIDGE_OP_HDMI_HDR_DRM_INFOFRAME |
 			   DRM_BRIDGE_OP_HDMI_SPD_INFOFRAME;
-	if (!hdmi->no_hpd)
-		hdmi->bridge.ops |= DRM_BRIDGE_OP_HPD;
 	hdmi->bridge.of_node = pdev->dev.of_node;
 	hdmi->bridge.type = DRM_MODE_CONNECTOR_HDMIA;
 	hdmi->bridge.vendor = "Synopsys";
 	hdmi->bridge.product = "DW HDMI QP TX";
 
+	if (hdmi->no_hpd) {
+		hdmi->bridge.supported_hdmi_ver = HDMI_VERSION_1_4;
+	} else {
+		hdmi->bridge.supported_hdmi_ver = HDMI_VERSION_2_0;
+		hdmi->bridge.ops |= DRM_BRIDGE_OP_HPD;
+	}
+
 	if (plat_data->supported_formats)
 		hdmi->bridge.supported_formats = plat_data->supported_formats;
 

-- 
2.55.0

[PATCH v12 05/16] drm/bridge: dw-hdmi-qp: Provide dw_hdmi_qp_hpd_notify() helper

[Cristian Ciocaltea]

Export dw_hdmi_qp_hpd_notify() for platform drivers to report hot-plug
detection events.  Unlike drm_helper_hpd_irq_event() which polls all
connectors, this helper targets only the affected connector and ensures
.detect_ctx() is invoked on reconnection events to trigger SCDC state
recovery.

Track the connector in use in curr_conn, set and cleared by the atomic
enable/disable hooks.  Since the helper runs asynchronously from
IRQ/work context, hold a reference on that connector for as long as the
bridge is enabled, so that the pointer is always safe to dereference.
Otherwise validity would rest on the platform glue quiescing HPD before
the DRM-managed connector is freed during teardown, risking a
use-after-free on a late HPD event.

The helper must take its own temporary reference while running, because
the long-lived reference above only guarantees the pointer is valid at
the time it is read, not that the connector survives a concurrent
atomic_disable dropping that reference while the helper is still using
it.  Moreover, reading the pointer and acquiring the temporary reference
must be atomic with respect to atomic_disable, otherwise a window exists
where drm_connector_get() is called on a connector whose refcount
already reached zero, hence guard curr_conn with a mutex for this
purpose.  Holding the mutex across the whole helper instead is not an
option, as it takes modeset locks while atomic_disable is called with
them held.

Also note that this helper may sleep, so it must only be called from a
threaded HPD interrupt handler or a work function.

Tested-by: Maud Spierings <maud_spierings@hotmail.com>
Tested-by: Diederik de Haas <diederik@cknow-tech.com>  # NanoPC-T6 LTS, Rock 5B
Tested-by: Robin Räber <robin.raebi@gmail.com>
Tested-by: Igor Paunovic <royalnet026@gmail.com> # RK3588 Orange Pi 5 Plus, dual HDMI
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
---
 drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c | 59 ++++++++++++++++++++++++++++
 include/drm/bridge/dw_hdmi_qp.h              |  1 +
 2 files changed, 60 insertions(+)

diff --git a/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c b/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c
index 14723ee5cb05..54181b052fe2 100644
--- a/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c
+++ b/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c
@@ -28,6 +28,7 @@
 #include <drm/drm_edid.h>
 #include <drm/drm_modes.h>
 #include <drm/drm_print.h>
+#include <drm/drm_probe_helper.h>
 
 #include <media/cec.h>
 
@@ -81,6 +82,9 @@ struct dw_hdmi_qp {
 	struct regmap *regm;
 	int main_irq;
 
+	/* Guards curr_conn between atomic hooks and HPD notify readers */
+	struct mutex mutex;
+	struct drm_connector *curr_conn;
 	unsigned long tmds_char_rate;
 	bool no_hpd;
 };
@@ -564,6 +568,10 @@ static void dw_hdmi_qp_bridge_atomic_enable(struct drm_bridge *bridge,
 		dev_dbg(hdmi->dev, "%s mode=DVI\n", __func__);
 	}
 
+	drm_connector_get(connector);
+	scoped_guard(mutex, &hdmi->mutex)
+		hdmi->curr_conn = connector;
+
 	hdmi->phy.ops->init(hdmi, hdmi->phy.data);
 
 	dw_hdmi_qp_mod(hdmi, HDCP2_BYPASS, HDCP2_BYPASS, HDCP2LOGIC_CONFIG0);
@@ -578,6 +586,13 @@ static void dw_hdmi_qp_bridge_atomic_disable(struct drm_bridge *bridge,
 	struct dw_hdmi_qp *hdmi = bridge->driver_private;
 	struct drm_connector *connector;
 
+	scoped_guard(mutex, &hdmi->mutex) {
+		if (hdmi->curr_conn) {
+			drm_connector_put(hdmi->curr_conn);
+			hdmi->curr_conn = NULL;
+		}
+	}
+
 	hdmi->tmds_char_rate = 0;
 
 	connector = drm_atomic_get_old_connector_for_encoder(state, bridge->encoder);
@@ -1099,6 +1114,10 @@ struct dw_hdmi_qp *dw_hdmi_qp_bind(struct platform_device *pdev,
 	if (IS_ERR(hdmi))
 		return ERR_CAST(hdmi);
 
+	ret = devm_mutex_init(dev, &hdmi->mutex);
+	if (ret)
+		return ERR_PTR(ret);
+
 	hdmi->dev = dev;
 
 	regs = devm_platform_ioremap_resource(pdev, 0);
@@ -1194,6 +1213,46 @@ struct dw_hdmi_qp *dw_hdmi_qp_bind(struct platform_device *pdev,
 }
 EXPORT_SYMBOL_GPL(dw_hdmi_qp_bind);
 
+/**
+ * dw_hdmi_qp_hpd_notify() - Notify a hot-plug detection event
+ * @hdmi: pointer to the DW HDMI QP controller
+ *
+ * Platform drivers should call this from their threaded IRQ handler or work
+ * function to notify the bridge of a connection status change.
+ * The bridge's .read_hpd() phy_ops callback is used to read the current
+ * connection status.
+ */
+void dw_hdmi_qp_hpd_notify(struct dw_hdmi_qp *hdmi)
+{
+	enum drm_connector_status status;
+	struct drm_connector *conn;
+
+	scoped_guard(mutex, &hdmi->mutex) {
+		conn = hdmi->curr_conn;
+		if (conn)
+			drm_connector_get(conn);
+	}
+
+	status = hdmi->phy.ops->read_hpd(hdmi, hdmi->phy.data);
+	dev_dbg(hdmi->dev, "%s status=%d\n", __func__, status);
+
+	/*
+	 * When the display pipeline has been already active, switch to
+	 * drm_connector_helper_hpd_irq_event() to ensure .detect_ctx()
+	 * gets invoked, i.e. via drm_helper_probe_detect(), because
+	 * drm_bridge_hpd_notify() defers to a delayed hotplug path in
+	 * this case.
+	 */
+	if (conn && status == connector_status_connected)
+		drm_connector_helper_hpd_irq_event(conn);
+	else
+		drm_bridge_hpd_notify(&hdmi->bridge, status);
+
+	if (conn)
+		drm_connector_put(conn);
+}
+EXPORT_SYMBOL_GPL(dw_hdmi_qp_hpd_notify);
+
 void dw_hdmi_qp_suspend(struct device *dev, struct dw_hdmi_qp *hdmi)
 {
 	disable_irq(hdmi->main_irq);
diff --git a/include/drm/bridge/dw_hdmi_qp.h b/include/drm/bridge/dw_hdmi_qp.h
index b80fceffc315..b4fb1c578a5b 100644
--- a/include/drm/bridge/dw_hdmi_qp.h
+++ b/include/drm/bridge/dw_hdmi_qp.h
@@ -36,6 +36,7 @@ struct dw_hdmi_qp_plat_data {
 struct dw_hdmi_qp *dw_hdmi_qp_bind(struct platform_device *pdev,
 				   struct drm_encoder *encoder,
 				   const struct dw_hdmi_qp_plat_data *plat_data);
+void dw_hdmi_qp_hpd_notify(struct dw_hdmi_qp *hdmi);
 void dw_hdmi_qp_suspend(struct device *dev, struct dw_hdmi_qp *hdmi);
 void dw_hdmi_qp_resume(struct device *dev, struct dw_hdmi_qp *hdmi);
 #endif /* __DW_HDMI_QP__ */

-- 
2.55.0

[PATCH v12 06/16] drm/rockchip: dw_hdmi_qp: Fix invalid drvdata access in PM ops

[Cristian Ciocaltea]

The system-sleep PM callbacks fetch the driver state via
dev_get_drvdata() and dereference it unconditionally.  However, the
platform device is bound by ->probe(), which merely registers a
component, hence the callbacks may well run when no valid state is
available:

- Before the aggregate driver calls ->bind(), e.g. when the system
  suspends while probing is still deferred waiting for the VOP,
  dev_get_drvdata() returns NULL.

- After ->unbind(), or when ->bind() fails past platform_set_drvdata(),
  drvdata is left pointing to memory obtained via drmm_kzalloc(), which
  is released together with the DRM device, while this platform device
  remains bound and PM-capable.

Publish the driver data only after ->bind() completes successfully, and
clear it in ->unbind().  This ensures that a non-NULL drvdata always
represents live state, allowing each PM callback to rely on a single
NULL check.

Fixes: 128a9bf8ace2 ("drm/rockchip: Add basic RK3588 HDMI output support")
Reported-by: Sashiko <sashiko-bot@kernel.org>
Closes: https://lore.kernel.org/all/20260702151550.8E14D1F000E9@smtp.kernel.org/
Closes: https://lore.kernel.org/all/20260723014811.1F9F01F000E9@smtp.kernel.org/
Tested-by: Robin Räber <robin.raebi@gmail.com>
Tested-by: Igor Paunovic <royalnet026@gmail.com> # RK3588 Orange Pi 5 Plus, dual HDMI
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
---
 drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c | 12 +++++++++---
 1 file changed, 9 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c b/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
index 9c4cf68d79d5..646a4cb96830 100644
--- a/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
+++ b/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
@@ -683,8 +683,6 @@ static int dw_hdmi_qp_rockchip_bind(struct device *dev, struct device *master,
 	if (ret)
 		return dev_err_probe(hdmi->dev, ret, "Failed to init encoder");
 
-	platform_set_drvdata(pdev, hdmi);
-
 	hdmi->hdmi = dw_hdmi_qp_bind(pdev, encoder, &plat_data);
 	if (IS_ERR(hdmi->hdmi))
 		return dev_err_probe(hdmi->dev, PTR_ERR(hdmi->hdmi),
@@ -695,6 +693,8 @@ static int dw_hdmi_qp_rockchip_bind(struct device *dev, struct device *master,
 		return dev_err_probe(hdmi->dev, PTR_ERR(connector),
 				     "Failed to init bridge connector\n");
 
+	platform_set_drvdata(pdev, hdmi);
+
 	return 0;
 }
 
@@ -704,6 +704,8 @@ static void dw_hdmi_qp_rockchip_unbind(struct device *dev,
 {
 	struct rockchip_hdmi_qp *hdmi = dev_get_drvdata(dev);
 
+	dev_set_drvdata(dev, NULL);
+
 	cancel_delayed_work_sync(&hdmi->hpd_work);
 }
 
@@ -726,7 +728,8 @@ static int __maybe_unused dw_hdmi_qp_rockchip_suspend(struct device *dev)
 {
 	struct rockchip_hdmi_qp *hdmi = dev_get_drvdata(dev);
 
-	dw_hdmi_qp_suspend(dev, hdmi->hdmi);
+	if (hdmi)
+		dw_hdmi_qp_suspend(dev, hdmi->hdmi);
 
 	return 0;
 }
@@ -735,6 +738,9 @@ static int __maybe_unused dw_hdmi_qp_rockchip_resume(struct device *dev)
 {
 	struct rockchip_hdmi_qp *hdmi = dev_get_drvdata(dev);
 
+	if (!hdmi)
+		return 0;
+
 	hdmi->ctrl_ops->io_init(hdmi);
 
 	dw_hdmi_qp_resume(dev, hdmi->hdmi);

-- 
2.55.0

[PATCH v12 07/16] drm/rockchip: dw_hdmi_qp: Cancel pending HPD work on suspend

[Cristian Ciocaltea]

The HPD threaded handler debounces hotplug events by delaying
hdmi->hpd_work for 150 ms.  Nothing discards that work when the system
goes to sleep: drm_kms_helper_poll_disable() only disables the HPD IRQ
line via the bridge .hpd_disable() op, and disable_irq() merely waits
for the threaded handler, which has already returned after arming the
timer.

If the debounce expires during suspend, its non-freezable work may probe
the connector and read the EDID over DDC.  Since genpd powers off the
HDMI controller's power domain in ->suspend_noirq(), this access may
hit unpowered hardware and hang the system.

Cancel the work in the suspend callback.  The DRM device is registered
as a consumer of the HDMI device, so it always suspends first and the
HPD IRQ is already disabled by that point, meaning nothing can re-arm
the work behind our back.  No counterpart is needed on resume, as the
work gets scheduled again on the next hotplug interrupt.

Fixes: 128a9bf8ace2 ("drm/rockchip: Add basic RK3588 HDMI output support")
Reported-by: Sashiko <sashiko-bot@kernel.org>
Closes: https://lore.kernel.org/all/20260723015909.7F79A1F000E9@smtp.kernel.org/
Tested-by: Robin Räber <robin.raebi@gmail.com>
Tested-by: Igor Paunovic <royalnet026@gmail.com> # RK3588 Orange Pi 5 Plus, dual HDMI
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
---
 drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c b/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
index 646a4cb96830..42f2627e1a05 100644
--- a/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
+++ b/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
@@ -728,8 +728,12 @@ static int __maybe_unused dw_hdmi_qp_rockchip_suspend(struct device *dev)
 {
 	struct rockchip_hdmi_qp *hdmi = dev_get_drvdata(dev);
 
-	if (hdmi)
-		dw_hdmi_qp_suspend(dev, hdmi->hdmi);
+	if (!hdmi)
+		return 0;
+
+	cancel_delayed_work_sync(&hdmi->hpd_work);
+
+	dw_hdmi_qp_suspend(dev, hdmi->hdmi);
 
 	return 0;
 }

-- 
2.55.0

[PATCH v12 08/16] drm/rockchip: dw_hdmi_qp: Add missing newlines in dev_err_probe() messages

[Cristian Ciocaltea]

Add the missing trailing newlines to a couple of dev_err_probe() calls
in dw_hdmi_qp_rockchip_bind().

Fixes: b6736a4ea3fa ("drm/rockchip: dw_hdmi_qp: Improve error handling with dev_err_probe()")
Fixes: e1f7b7cbd74c ("drm/rockchip: dw_hdmi_qp: Switch to drmm_encoder_init()")
Reviewed-by: Heiko Stuebner <heiko@sntech.de>
Tested-by: Maud Spierings <maud_spierings@hotmail.com>
Tested-by: Diederik de Haas <diederik@cknow-tech.com>  # NanoPC-T6 LTS, Rock 5B
Tested-by: Robin Räber <robin.raebi@gmail.com>
Tested-by: Igor Paunovic <royalnet026@gmail.com> # RK3588 Orange Pi 5 Plus, dual HDMI
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
---
 drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c b/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
index 42f2627e1a05..df0e2bb1eedc 100644
--- a/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
+++ b/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
@@ -681,12 +681,12 @@ static int dw_hdmi_qp_rockchip_bind(struct device *dev, struct device *master,
 	drm_encoder_helper_add(encoder, &dw_hdmi_qp_rockchip_encoder_helper_funcs);
 	ret = drmm_encoder_init(drm, encoder, NULL, DRM_MODE_ENCODER_TMDS, NULL);
 	if (ret)
-		return dev_err_probe(hdmi->dev, ret, "Failed to init encoder");
+		return dev_err_probe(hdmi->dev, ret, "Failed to init encoder\n");
 
 	hdmi->hdmi = dw_hdmi_qp_bind(pdev, encoder, &plat_data);
 	if (IS_ERR(hdmi->hdmi))
 		return dev_err_probe(hdmi->dev, PTR_ERR(hdmi->hdmi),
-				     "Failed to bind dw-hdmi-qp");
+				     "Failed to bind dw-hdmi-qp\n");
 
 	connector = drm_bridge_connector_init(drm, encoder);
 	if (IS_ERR(connector))

-- 
2.55.0

[PATCH v12 09/16] drm/rockchip: dw_hdmi_qp: Use local dev variable consistently in bind()

[Cristian Ciocaltea]

Replace indirect struct device accesses via hdmi->dev and pdev->dev with
the local dev parameter already available in dw_hdmi_qp_rockchip_bind(),
for consistency and readability.

Reviewed-by: Heiko Stuebner <heiko@sntech.de>
Tested-by: Maud Spierings <maud_spierings@hotmail.com>
Tested-by: Diederik de Haas <diederik@cknow-tech.com>  # NanoPC-T6 LTS, Rock 5B
Tested-by: Robin Räber <robin.raebi@gmail.com>
Tested-by: Igor Paunovic <royalnet026@gmail.com> # RK3588 Orange Pi 5 Plus, dual HDMI
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
---
 drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c | 33 +++++++++++++-------------
 1 file changed, 16 insertions(+), 17 deletions(-)

diff --git a/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c b/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
index df0e2bb1eedc..456bf2f194d7 100644
--- a/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
+++ b/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
@@ -563,7 +563,7 @@ static int dw_hdmi_qp_rockchip_bind(struct device *dev, struct device *master,
 	struct clk *ref_clk;
 	int ret, irq, i;
 
-	if (!pdev->dev.of_node)
+	if (!dev->of_node)
 		return -ENODEV;
 
 	hdmi = drmm_kzalloc(drm, sizeof(*hdmi), GFP_KERNEL);
@@ -583,7 +583,7 @@ static int dw_hdmi_qp_rockchip_bind(struct device *dev, struct device *master,
 		return dev_err_probe(dev, -ENODEV, "Missing platform ctrl ops\n");
 
 	hdmi->ctrl_ops = cfg->ctrl_ops;
-	hdmi->dev = &pdev->dev;
+	hdmi->dev = dev;
 	hdmi->port_id = -ENODEV;
 
 	/* Identify port ID by matching base IO address */
@@ -594,7 +594,7 @@ static int dw_hdmi_qp_rockchip_bind(struct device *dev, struct device *master,
 		}
 	}
 	if (hdmi->port_id < 0)
-		return dev_err_probe(hdmi->dev, hdmi->port_id,
+		return dev_err_probe(dev, hdmi->port_id,
 				     "Failed to match HDMI port ID\n");
 
 	plat_data.phy_ops = cfg->phy_ops;
@@ -622,37 +622,36 @@ static int dw_hdmi_qp_rockchip_bind(struct device *dev, struct device *master,
 	hdmi->regmap = syscon_regmap_lookup_by_phandle(dev->of_node,
 						       "rockchip,grf");
 	if (IS_ERR(hdmi->regmap))
-		return dev_err_probe(hdmi->dev, PTR_ERR(hdmi->regmap),
+		return dev_err_probe(dev, PTR_ERR(hdmi->regmap),
 				     "Unable to get rockchip,grf\n");
 
 	hdmi->vo_regmap = syscon_regmap_lookup_by_phandle(dev->of_node,
 							  "rockchip,vo-grf");
 	if (IS_ERR(hdmi->vo_regmap))
-		return dev_err_probe(hdmi->dev, PTR_ERR(hdmi->vo_regmap),
+		return dev_err_probe(dev, PTR_ERR(hdmi->vo_regmap),
 				     "Unable to get rockchip,vo-grf\n");
 
-	ret = devm_clk_bulk_get_all_enabled(hdmi->dev, &clks);
+	ret = devm_clk_bulk_get_all_enabled(dev, &clks);
 	if (ret < 0)
-		return dev_err_probe(hdmi->dev, ret, "Failed to get clocks\n");
+		return dev_err_probe(dev, ret, "Failed to get clocks\n");
 
-	ref_clk = clk_get(hdmi->dev, "ref");
+	ref_clk = clk_get(dev, "ref");
 	if (IS_ERR(ref_clk))
-		return dev_err_probe(hdmi->dev, PTR_ERR(ref_clk),
+		return dev_err_probe(dev, PTR_ERR(ref_clk),
 				     "Failed to get ref clock\n");
 
 	plat_data.ref_clk_rate = clk_get_rate(ref_clk);
 	clk_put(ref_clk);
 
-	hdmi->frl_enable_gpio = devm_gpiod_get_optional(hdmi->dev, "frl-enable",
+	hdmi->frl_enable_gpio = devm_gpiod_get_optional(dev, "frl-enable",
 							GPIOD_OUT_LOW);
 	if (IS_ERR(hdmi->frl_enable_gpio))
-		return dev_err_probe(hdmi->dev, PTR_ERR(hdmi->frl_enable_gpio),
+		return dev_err_probe(dev, PTR_ERR(hdmi->frl_enable_gpio),
 				     "Failed to request FRL enable GPIO\n");
 
 	hdmi->phy = devm_of_phy_get_by_index(dev, dev->of_node, 0);
 	if (IS_ERR(hdmi->phy))
-		return dev_err_probe(hdmi->dev, PTR_ERR(hdmi->phy),
-				     "Failed to get phy\n");
+		return dev_err_probe(dev, PTR_ERR(hdmi->phy), "Failed to get phy\n");
 
 	cfg->ctrl_ops->io_init(hdmi);
 
@@ -670,7 +669,7 @@ static int dw_hdmi_qp_rockchip_bind(struct device *dev, struct device *master,
 	if (irq < 0)
 		return irq;
 
-	ret = devm_request_threaded_irq(hdmi->dev, irq,
+	ret = devm_request_threaded_irq(dev, irq,
 					cfg->ctrl_ops->hardirq_callback,
 					cfg->ctrl_ops->irq_callback,
 					IRQF_SHARED, "dw-hdmi-qp-hpd",
@@ -681,16 +680,16 @@ static int dw_hdmi_qp_rockchip_bind(struct device *dev, struct device *master,
 	drm_encoder_helper_add(encoder, &dw_hdmi_qp_rockchip_encoder_helper_funcs);
 	ret = drmm_encoder_init(drm, encoder, NULL, DRM_MODE_ENCODER_TMDS, NULL);
 	if (ret)
-		return dev_err_probe(hdmi->dev, ret, "Failed to init encoder\n");
+		return dev_err_probe(dev, ret, "Failed to init encoder\n");
 
 	hdmi->hdmi = dw_hdmi_qp_bind(pdev, encoder, &plat_data);
 	if (IS_ERR(hdmi->hdmi))
-		return dev_err_probe(hdmi->dev, PTR_ERR(hdmi->hdmi),
+		return dev_err_probe(dev, PTR_ERR(hdmi->hdmi),
 				     "Failed to bind dw-hdmi-qp\n");
 
 	connector = drm_bridge_connector_init(drm, encoder);
 	if (IS_ERR(connector))
-		return dev_err_probe(hdmi->dev, PTR_ERR(connector),
+		return dev_err_probe(dev, PTR_ERR(connector),
 				     "Failed to init bridge connector\n");
 
 	platform_set_drvdata(pdev, hdmi);

-- 
2.55.0

[PATCH v12 10/16] drm/rockchip: dw_hdmi_qp: Avoid spurious HPD IRQ thread wakeups

[Cristian Ciocaltea]

The RK3576 and RK3588 HPD hardirq handlers read shared status registers
and currently wake the threaded handler whenever any bit in those
registers is set.  This can wake the HDMI IRQ thread for unrelated
status bits, and on RK3588 can also wake the thread for the other HDMI
port.

Mask the status value with the SoC-specific HPD interrupt bit before
deciding whether to wake the threaded handler.  On RK3588, select the
HDMI0 or HDMI1 HPD bit according to the instance port id.

No stable backport is needed.  The issue may only cause spurious IRQ
thread wakeups and does not cause functional breakage.

Reported-by: Sashiko <sashiko-bot@kernel.org>
Closes: https://lore.kernel.org/all/20260702151346.5A4EC1F000E9@smtp.kernel.org/
Tested-by: Diederik de Haas <diederik@cknow-tech.com>  # NanoPC-T6 LTS, Rock 5B
Tested-by: Robin Räber <robin.raebi@gmail.com>
Tested-by: Igor Paunovic <royalnet026@gmail.com> # RK3588 Orange Pi 5 Plus, dual HDMI
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
---
 drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c b/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
index 456bf2f194d7..dff400233154 100644
--- a/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
+++ b/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
@@ -34,6 +34,7 @@
 
 #define RK3576_IOC_HDMI_HPD_STATUS	0xa440
 #define RK3576_HDMI_LEVEL_INT		BIT(3)
+#define RK3576_HDMI_OHPD_INT		BIT(4)
 
 #define RK3576_VO0_GRF_SOC_CON1		0x0004
 #define RK3576_HDMI_FRL_MOD		BIT(0)
@@ -64,7 +65,9 @@
 #define RK3588_HPD_HDMI1_IO_EN_MASK	BIT(13)
 #define RK3588_GRF_SOC_STATUS1		0x0384
 #define RK3588_HDMI0_LEVEL_INT		BIT(16)
+#define RK3588_HDMI0_OHPD_INT		BIT(17)
 #define RK3588_HDMI1_LEVEL_INT		BIT(24)
+#define RK3588_HDMI1_OHPD_INT		BIT(25)
 #define RK3588_GRF_VO1_CON3		0x000c
 #define RK3588_GRF_VO1_CON6		0x0018
 #define RK3588_COLOR_DEPTH_MASK		GENMASK(7, 4)
@@ -348,7 +351,7 @@ static irqreturn_t dw_hdmi_qp_rk3576_hardirq(int irq, void *dev_id)
 	u32 intr_stat, val;
 
 	regmap_read(hdmi->regmap, RK3576_IOC_HDMI_HPD_STATUS, &intr_stat);
-	if (intr_stat) {
+	if (intr_stat & RK3576_HDMI_OHPD_INT) {
 		val = FIELD_PREP_WM16(RK3576_HDMI_HPD_INT_MSK, 1);
 
 		regmap_write(hdmi->regmap, RK3576_IOC_MISC_CON0, val);
@@ -381,6 +384,8 @@ static irqreturn_t dw_hdmi_qp_rk3588_hardirq(int irq, void *dev_id)
 
 	regmap_read(hdmi->regmap, RK3588_GRF_SOC_STATUS1, &intr_stat);
 
+	intr_stat &= hdmi->port_id ? RK3588_HDMI1_OHPD_INT : RK3588_HDMI0_OHPD_INT;
+
 	if (intr_stat) {
 		if (hdmi->port_id)
 			val = FIELD_PREP_WM16(RK3588_HDMI1_HPD_INT_MSK, 1);

-- 
2.55.0

[PATCH v12 11/16] drm/rockchip: dw_hdmi_qp: Mask RK3576 HPD IRQ in io_init

[Cristian Ciocaltea]

Start with HPD masked, matching RK3588, so the bridge framework controls
when interrupts are armed via hpd_enable().

Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
---
 drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c b/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
index dff400233154..bcfa6646546e 100644
--- a/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
+++ b/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
@@ -432,7 +432,7 @@ static void dw_hdmi_qp_rk3576_io_init(struct rockchip_hdmi_qp *hdmi)
 
 	regmap_write(hdmi->vo_regmap, RK3576_VO0_GRF_SOC_CON14, val);
 
-	val = FIELD_PREP_WM16(RK3576_HDMI_HPD_INT_MSK, 0);
+	val = FIELD_PREP_WM16(RK3576_HDMI_HPD_INT_MSK, 1);
 	regmap_write(hdmi->regmap, RK3576_IOC_MISC_CON0, val);
 }
 

-- 
2.55.0

[PATCH v12 12/16] drm/rockchip: dw_hdmi_qp: Implement .{enable,disable}_hpd() PHY ops

[Cristian Ciocaltea]

Implement .enable_hpd() and .disable_hpd() phy ops for RK3576 and RK3588
SoCs, used by the corresponding bridge callbacks for HPD activation
control.

Consolidate the interrupt clear-and-unmask sequence into enable_hpd()
and the mask-only operation into disable_hpd(), replacing the open-coded
register writes in the interrupt handlers and io_init().

The .setup_hpd() phy op, which was previously called from
dw_hdmi_qp_init_hw(), is no longer needed and its users are removed.

Tested-by: Maud Spierings <maud_spierings@hotmail.com>
Tested-by: Diederik de Haas <diederik@cknow-tech.com>  # NanoPC-T6 LTS, Rock 5B
Tested-by: Robin Räber <robin.raebi@gmail.com>
Tested-by: Igor Paunovic <royalnet026@gmail.com> # RK3588 Orange Pi 5 Plus, dual HDMI
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
---
 drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c | 78 +++++++++++++-------------
 1 file changed, 38 insertions(+), 40 deletions(-)

diff --git a/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c b/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
index bcfa6646546e..9e48ec06eb66 100644
--- a/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
+++ b/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
@@ -277,7 +277,7 @@ dw_hdmi_qp_rk3588_read_hpd(struct dw_hdmi_qp *dw_hdmi, void *data)
 	return val ? connector_status_connected : connector_status_disconnected;
 }
 
-static void dw_hdmi_qp_rk3588_setup_hpd(struct dw_hdmi_qp *dw_hdmi, void *data)
+static void dw_hdmi_qp_rk3588_enable_hpd(struct dw_hdmi_qp *dw_hdmi, void *data)
 {
 	struct rockchip_hdmi_qp *hdmi = (struct rockchip_hdmi_qp *)data;
 	u32 val;
@@ -292,11 +292,25 @@ static void dw_hdmi_qp_rk3588_setup_hpd(struct dw_hdmi_qp *dw_hdmi, void *data)
 	regmap_write(hdmi->regmap, RK3588_GRF_SOC_CON2, val);
 }
 
+static void dw_hdmi_qp_rk3588_disable_hpd(struct dw_hdmi_qp *dw_hdmi, void *data)
+{
+	struct rockchip_hdmi_qp *hdmi = (struct rockchip_hdmi_qp *)data;
+	u32 val;
+
+	if (hdmi->port_id)
+		val = FIELD_PREP_WM16(RK3588_HDMI1_HPD_INT_MSK, 1);
+	else
+		val = FIELD_PREP_WM16(RK3588_HDMI0_HPD_INT_MSK, 1);
+
+	regmap_write(hdmi->regmap, RK3588_GRF_SOC_CON2, val);
+}
+
 static const struct dw_hdmi_qp_phy_ops rk3588_hdmi_phy_ops = {
 	.init		= dw_hdmi_qp_rk3588_phy_init,
 	.disable	= dw_hdmi_qp_rk3588_phy_disable,
 	.read_hpd	= dw_hdmi_qp_rk3588_read_hpd,
-	.setup_hpd	= dw_hdmi_qp_rk3588_setup_hpd,
+	.enable_hpd	= dw_hdmi_qp_rk3588_enable_hpd,
+	.disable_hpd	= dw_hdmi_qp_rk3588_disable_hpd,
 };
 
 static enum drm_connector_status
@@ -311,7 +325,7 @@ dw_hdmi_qp_rk3576_read_hpd(struct dw_hdmi_qp *dw_hdmi, void *data)
 		connector_status_connected : connector_status_disconnected;
 }
 
-static void dw_hdmi_qp_rk3576_setup_hpd(struct dw_hdmi_qp *dw_hdmi, void *data)
+static void dw_hdmi_qp_rk3576_enable_hpd(struct dw_hdmi_qp *dw_hdmi, void *data)
 {
 	struct rockchip_hdmi_qp *hdmi = (struct rockchip_hdmi_qp *)data;
 	u32 val;
@@ -320,14 +334,22 @@ static void dw_hdmi_qp_rk3576_setup_hpd(struct dw_hdmi_qp *dw_hdmi, void *data)
 	       FIELD_PREP_WM16(RK3576_HDMI_HPD_INT_MSK, 0));
 
 	regmap_write(hdmi->regmap, RK3576_IOC_MISC_CON0, val);
-	regmap_write(hdmi->regmap, 0xa404, 0xffff0102);
+}
+
+static void dw_hdmi_qp_rk3576_disable_hpd(struct dw_hdmi_qp *dw_hdmi, void *data)
+{
+	struct rockchip_hdmi_qp *hdmi = (struct rockchip_hdmi_qp *)data;
+
+	regmap_write(hdmi->regmap, RK3576_IOC_MISC_CON0,
+		     FIELD_PREP_WM16(RK3576_HDMI_HPD_INT_MSK, 1));
 }
 
 static const struct dw_hdmi_qp_phy_ops rk3576_hdmi_phy_ops = {
 	.init		= dw_hdmi_qp_rk3588_phy_init,
 	.disable	= dw_hdmi_qp_rk3588_phy_disable,
 	.read_hpd	= dw_hdmi_qp_rk3576_read_hpd,
-	.setup_hpd	= dw_hdmi_qp_rk3576_setup_hpd,
+	.enable_hpd	= dw_hdmi_qp_rk3576_enable_hpd,
+	.disable_hpd	= dw_hdmi_qp_rk3576_disable_hpd,
 };
 
 static void dw_hdmi_qp_rk3588_hpd_work(struct work_struct *work)
@@ -348,13 +370,12 @@ static void dw_hdmi_qp_rk3588_hpd_work(struct work_struct *work)
 static irqreturn_t dw_hdmi_qp_rk3576_hardirq(int irq, void *dev_id)
 {
 	struct rockchip_hdmi_qp *hdmi = dev_id;
-	u32 intr_stat, val;
+	u32 intr_stat;
 
 	regmap_read(hdmi->regmap, RK3576_IOC_HDMI_HPD_STATUS, &intr_stat);
-	if (intr_stat & RK3576_HDMI_OHPD_INT) {
-		val = FIELD_PREP_WM16(RK3576_HDMI_HPD_INT_MSK, 1);
 
-		regmap_write(hdmi->regmap, RK3576_IOC_MISC_CON0, val);
+	if (intr_stat & RK3576_HDMI_OHPD_INT) {
+		dw_hdmi_qp_rk3576_disable_hpd(NULL, hdmi);
 		return IRQ_WAKE_THREAD;
 	}
 
@@ -364,15 +385,11 @@ static irqreturn_t dw_hdmi_qp_rk3576_hardirq(int irq, void *dev_id)
 static irqreturn_t dw_hdmi_qp_rk3576_irq(int irq, void *dev_id)
 {
 	struct rockchip_hdmi_qp *hdmi = dev_id;
-	u32 val;
 
-	val = FIELD_PREP_WM16(RK3576_HDMI_HPD_INT_CLR, 1);
-	regmap_write(hdmi->regmap, RK3576_IOC_MISC_CON0, val);
 	mod_delayed_work(system_percpu_wq, &hdmi->hpd_work,
 			 msecs_to_jiffies(HOTPLUG_DEBOUNCE_MS));
 
-	val = FIELD_PREP_WM16(RK3576_HDMI_HPD_INT_MSK, 0);
-	regmap_write(hdmi->regmap, RK3576_IOC_MISC_CON0, val);
+	dw_hdmi_qp_rk3576_enable_hpd(NULL, hdmi);
 
 	return IRQ_HANDLED;
 }
@@ -380,18 +397,14 @@ static irqreturn_t dw_hdmi_qp_rk3576_irq(int irq, void *dev_id)
 static irqreturn_t dw_hdmi_qp_rk3588_hardirq(int irq, void *dev_id)
 {
 	struct rockchip_hdmi_qp *hdmi = dev_id;
-	u32 intr_stat, val;
+	u32 intr_stat;
 
 	regmap_read(hdmi->regmap, RK3588_GRF_SOC_STATUS1, &intr_stat);
 
 	intr_stat &= hdmi->port_id ? RK3588_HDMI1_OHPD_INT : RK3588_HDMI0_OHPD_INT;
 
 	if (intr_stat) {
-		if (hdmi->port_id)
-			val = FIELD_PREP_WM16(RK3588_HDMI1_HPD_INT_MSK, 1);
-		else
-			val = FIELD_PREP_WM16(RK3588_HDMI0_HPD_INT_MSK, 1);
-		regmap_write(hdmi->regmap, RK3588_GRF_SOC_CON2, val);
+		dw_hdmi_qp_rk3588_disable_hpd(NULL, hdmi);
 		return IRQ_WAKE_THREAD;
 	}
 
@@ -401,22 +414,11 @@ static irqreturn_t dw_hdmi_qp_rk3588_hardirq(int irq, void *dev_id)
 static irqreturn_t dw_hdmi_qp_rk3588_irq(int irq, void *dev_id)
 {
 	struct rockchip_hdmi_qp *hdmi = dev_id;
-	u32 val;
-
-	if (hdmi->port_id)
-		val = FIELD_PREP_WM16(RK3588_HDMI1_HPD_INT_CLR, 1);
-	else
-		val = FIELD_PREP_WM16(RK3588_HDMI0_HPD_INT_CLR, 1);
-	regmap_write(hdmi->regmap, RK3588_GRF_SOC_CON2, val);
 
 	mod_delayed_work(system_percpu_wq, &hdmi->hpd_work,
 			 msecs_to_jiffies(HOTPLUG_DEBOUNCE_MS));
 
-	if (hdmi->port_id)
-		val |= FIELD_PREP_WM16(RK3588_HDMI1_HPD_INT_MSK, 0);
-	else
-		val |= FIELD_PREP_WM16(RK3588_HDMI0_HPD_INT_MSK, 0);
-	regmap_write(hdmi->regmap, RK3588_GRF_SOC_CON2, val);
+	dw_hdmi_qp_rk3588_enable_hpd(NULL, hdmi);
 
 	return IRQ_HANDLED;
 }
@@ -429,11 +431,11 @@ static void dw_hdmi_qp_rk3576_io_init(struct rockchip_hdmi_qp *hdmi)
 	      FIELD_PREP_WM16(RK3576_SDAIN_MASK, 1) |
 	      FIELD_PREP_WM16(RK3576_HDMI_GRANT_SEL, 1) |
 	      FIELD_PREP_WM16(RK3576_I2S_SEL_MASK, 1);
-
 	regmap_write(hdmi->vo_regmap, RK3576_VO0_GRF_SOC_CON14, val);
 
-	val = FIELD_PREP_WM16(RK3576_HDMI_HPD_INT_MSK, 1);
-	regmap_write(hdmi->regmap, RK3576_IOC_MISC_CON0, val);
+	regmap_write(hdmi->regmap, 0xa404, 0xffff0102);
+
+	dw_hdmi_qp_rk3576_disable_hpd(NULL, hdmi);
 }
 
 static void dw_hdmi_qp_rk3588_io_init(struct rockchip_hdmi_qp *hdmi)
@@ -458,11 +460,7 @@ static void dw_hdmi_qp_rk3588_io_init(struct rockchip_hdmi_qp *hdmi)
 		val = FIELD_PREP_WM16(RK3588_HDMI0_GRANT_SEL, 1);
 	regmap_write(hdmi->vo_regmap, RK3588_GRF_VO1_CON9, val);
 
-	if (hdmi->port_id)
-		val = FIELD_PREP_WM16(RK3588_HDMI1_HPD_INT_MSK, 1);
-	else
-		val = FIELD_PREP_WM16(RK3588_HDMI0_HPD_INT_MSK, 1);
-	regmap_write(hdmi->regmap, RK3588_GRF_SOC_CON2, val);
+	dw_hdmi_qp_rk3588_disable_hpd(NULL, hdmi);
 }
 
 static void dw_hdmi_qp_rk3576_enc_init(struct rockchip_hdmi_qp *hdmi,

-- 
2.55.0

[PATCH v12 13/16] drm/rockchip: dw_hdmi_qp: Factor out HPD interrupt (un)mask helpers

[Cristian Ciocaltea]

Extract the GRF-level HPD interrupt mask/unmask register writes into
dedicated dw_hdmi_qp_rk35{88,76}_{mask,unmask}_hpd() helpers, and use
them from the .enable_hpd()/.disable_hpd() phy ops as well as the
hardirq, threaded re-arm and io_init() paths.

This is a preparatory refactor that separates the SoC-level HPD mask
from the phy op entry points, so a subsequent change can gate the
platform HPD IRQ line without affecting the debounce paths.

Tested-by: Diederik de Haas <diederik@cknow-tech.com>  # NanoPC-T6 LTS, Rock 5B
Tested-by: Robin Räber <robin.raebi@gmail.com>
Tested-by: Igor Paunovic <royalnet026@gmail.com> # RK3588 Orange Pi 5 Plus, dual HDMI
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
---
 drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c | 54 ++++++++++++++++++--------
 1 file changed, 37 insertions(+), 17 deletions(-)

diff --git a/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c b/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
index 9e48ec06eb66..ecf12403101d 100644
--- a/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
+++ b/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
@@ -277,9 +277,8 @@ dw_hdmi_qp_rk3588_read_hpd(struct dw_hdmi_qp *dw_hdmi, void *data)
 	return val ? connector_status_connected : connector_status_disconnected;
 }
 
-static void dw_hdmi_qp_rk3588_enable_hpd(struct dw_hdmi_qp *dw_hdmi, void *data)
+static void dw_hdmi_qp_rk3588_unmask_hpd(struct rockchip_hdmi_qp *hdmi)
 {
-	struct rockchip_hdmi_qp *hdmi = (struct rockchip_hdmi_qp *)data;
 	u32 val;
 
 	if (hdmi->port_id)
@@ -292,9 +291,8 @@ static void dw_hdmi_qp_rk3588_enable_hpd(struct dw_hdmi_qp *dw_hdmi, void *data)
 	regmap_write(hdmi->regmap, RK3588_GRF_SOC_CON2, val);
 }
 
-static void dw_hdmi_qp_rk3588_disable_hpd(struct dw_hdmi_qp *dw_hdmi, void *data)
+static void dw_hdmi_qp_rk3588_mask_hpd(struct rockchip_hdmi_qp *hdmi)
 {
-	struct rockchip_hdmi_qp *hdmi = (struct rockchip_hdmi_qp *)data;
 	u32 val;
 
 	if (hdmi->port_id)
@@ -305,6 +303,20 @@ static void dw_hdmi_qp_rk3588_disable_hpd(struct dw_hdmi_qp *dw_hdmi, void *data
 	regmap_write(hdmi->regmap, RK3588_GRF_SOC_CON2, val);
 }
 
+static void dw_hdmi_qp_rk3588_enable_hpd(struct dw_hdmi_qp *dw_hdmi, void *data)
+{
+	struct rockchip_hdmi_qp *hdmi = (struct rockchip_hdmi_qp *)data;
+
+	dw_hdmi_qp_rk3588_unmask_hpd(hdmi);
+}
+
+static void dw_hdmi_qp_rk3588_disable_hpd(struct dw_hdmi_qp *dw_hdmi, void *data)
+{
+	struct rockchip_hdmi_qp *hdmi = (struct rockchip_hdmi_qp *)data;
+
+	dw_hdmi_qp_rk3588_mask_hpd(hdmi);
+}
+
 static const struct dw_hdmi_qp_phy_ops rk3588_hdmi_phy_ops = {
 	.init		= dw_hdmi_qp_rk3588_phy_init,
 	.disable	= dw_hdmi_qp_rk3588_phy_disable,
@@ -325,23 +337,31 @@ dw_hdmi_qp_rk3576_read_hpd(struct dw_hdmi_qp *dw_hdmi, void *data)
 		connector_status_connected : connector_status_disconnected;
 }
 
+static void dw_hdmi_qp_rk3576_unmask_hpd(struct rockchip_hdmi_qp *hdmi)
+{
+	regmap_write(hdmi->regmap, RK3576_IOC_MISC_CON0,
+		     FIELD_PREP_WM16(RK3576_HDMI_HPD_INT_CLR, 1) |
+		     FIELD_PREP_WM16(RK3576_HDMI_HPD_INT_MSK, 0));
+}
+
+static void dw_hdmi_qp_rk3576_mask_hpd(struct rockchip_hdmi_qp *hdmi)
+{
+	regmap_write(hdmi->regmap, RK3576_IOC_MISC_CON0,
+		     FIELD_PREP_WM16(RK3576_HDMI_HPD_INT_MSK, 1));
+}
+
 static void dw_hdmi_qp_rk3576_enable_hpd(struct dw_hdmi_qp *dw_hdmi, void *data)
 {
 	struct rockchip_hdmi_qp *hdmi = (struct rockchip_hdmi_qp *)data;
-	u32 val;
 
-	val = (FIELD_PREP_WM16(RK3576_HDMI_HPD_INT_CLR, 1) |
-	       FIELD_PREP_WM16(RK3576_HDMI_HPD_INT_MSK, 0));
-
-	regmap_write(hdmi->regmap, RK3576_IOC_MISC_CON0, val);
+	dw_hdmi_qp_rk3576_unmask_hpd(hdmi);
 }
 
 static void dw_hdmi_qp_rk3576_disable_hpd(struct dw_hdmi_qp *dw_hdmi, void *data)
 {
 	struct rockchip_hdmi_qp *hdmi = (struct rockchip_hdmi_qp *)data;
 
-	regmap_write(hdmi->regmap, RK3576_IOC_MISC_CON0,
-		     FIELD_PREP_WM16(RK3576_HDMI_HPD_INT_MSK, 1));
+	dw_hdmi_qp_rk3576_mask_hpd(hdmi);
 }
 
 static const struct dw_hdmi_qp_phy_ops rk3576_hdmi_phy_ops = {
@@ -375,7 +395,7 @@ static irqreturn_t dw_hdmi_qp_rk3576_hardirq(int irq, void *dev_id)
 	regmap_read(hdmi->regmap, RK3576_IOC_HDMI_HPD_STATUS, &intr_stat);
 
 	if (intr_stat & RK3576_HDMI_OHPD_INT) {
-		dw_hdmi_qp_rk3576_disable_hpd(NULL, hdmi);
+		dw_hdmi_qp_rk3576_mask_hpd(hdmi);
 		return IRQ_WAKE_THREAD;
 	}
 
@@ -389,7 +409,7 @@ static irqreturn_t dw_hdmi_qp_rk3576_irq(int irq, void *dev_id)
 	mod_delayed_work(system_percpu_wq, &hdmi->hpd_work,
 			 msecs_to_jiffies(HOTPLUG_DEBOUNCE_MS));
 
-	dw_hdmi_qp_rk3576_enable_hpd(NULL, hdmi);
+	dw_hdmi_qp_rk3576_unmask_hpd(hdmi);
 
 	return IRQ_HANDLED;
 }
@@ -404,7 +424,7 @@ static irqreturn_t dw_hdmi_qp_rk3588_hardirq(int irq, void *dev_id)
 	intr_stat &= hdmi->port_id ? RK3588_HDMI1_OHPD_INT : RK3588_HDMI0_OHPD_INT;
 
 	if (intr_stat) {
-		dw_hdmi_qp_rk3588_disable_hpd(NULL, hdmi);
+		dw_hdmi_qp_rk3588_mask_hpd(hdmi);
 		return IRQ_WAKE_THREAD;
 	}
 
@@ -418,7 +438,7 @@ static irqreturn_t dw_hdmi_qp_rk3588_irq(int irq, void *dev_id)
 	mod_delayed_work(system_percpu_wq, &hdmi->hpd_work,
 			 msecs_to_jiffies(HOTPLUG_DEBOUNCE_MS));
 
-	dw_hdmi_qp_rk3588_enable_hpd(NULL, hdmi);
+	dw_hdmi_qp_rk3588_unmask_hpd(hdmi);
 
 	return IRQ_HANDLED;
 }
@@ -435,7 +455,7 @@ static void dw_hdmi_qp_rk3576_io_init(struct rockchip_hdmi_qp *hdmi)
 
 	regmap_write(hdmi->regmap, 0xa404, 0xffff0102);
 
-	dw_hdmi_qp_rk3576_disable_hpd(NULL, hdmi);
+	dw_hdmi_qp_rk3576_mask_hpd(hdmi);
 }
 
 static void dw_hdmi_qp_rk3588_io_init(struct rockchip_hdmi_qp *hdmi)
@@ -460,7 +480,7 @@ static void dw_hdmi_qp_rk3588_io_init(struct rockchip_hdmi_qp *hdmi)
 		val = FIELD_PREP_WM16(RK3588_HDMI0_GRANT_SEL, 1);
 	regmap_write(hdmi->vo_regmap, RK3588_GRF_VO1_CON9, val);
 
-	dw_hdmi_qp_rk3588_disable_hpd(NULL, hdmi);
+	dw_hdmi_qp_rk3588_mask_hpd(hdmi);
 }
 
 static void dw_hdmi_qp_rk3576_enc_init(struct rockchip_hdmi_qp *hdmi,

-- 
2.55.0

[PATCH v12 14/16] drm/rockchip: dw_hdmi_qp: Control the HPD IRQ line via the bridge HPD ops

[Cristian Ciocaltea]

The HPD interrupt must not be delivered before the DRM connector is fully
set up, and it should be quiesced while the display pipeline is suspended.
On top of that, the threaded handler re-arms HPD unconditionally after the
hotplug debounce, which must not override an explicit disable requested by
the DRM core.

Request the HPD IRQ in bind() with IRQF_NO_AUTOEN so it stays disabled
until the core arms it, and drive enable_irq()/disable_irq() from the
phy ops invoked by the bridge through its hpd_{enable,disable}()
callbacks.  This ensures the line is only armed after the connector is
ready.  disable_irq() additionally waits for the threaded handler to
finish, so its GRF re-arm can no longer race with and override a
core-requested disable.

Masking the line is not enough on its own, though: the threaded handler
merely queues hpd_work with a 150 ms debounce, so an event taken shortly
before the disable would still probe the connector and read the EDID
long after the core considers HPD quiesced.  Since disable_irq() has
drained the threaded handler, which is the only site arming the work, it
can no longer be requeued and dropping it is race-free.

Use the non-syncing cancel_delayed_work() for that: waiting on a running
callback would pull mode_config.mutex, taken by the work through
drm_connector_helper_hpd_irq_event(), under the hpd_state_mutex held by
drm_bridge_hpd_disable().  That inverts the order of the connector probe
path, which holds mode_config.mutex while reaching hpd_state_mutex via
drm_kms_helper_poll_enable().  A callback already in progress is drained
by the system-sleep and unbind paths instead.

Note that IRQF_NO_AUTOEN is incompatible with IRQF_SHARED; the latter is
dropped as this IRQ has no other users.

Tested-by: Diederik de Haas <diederik@cknow-tech.com>  # NanoPC-T6 LTS, Rock 5B
Tested-by: Robin Räber <robin.raebi@gmail.com>
Tested-by: Igor Paunovic <royalnet026@gmail.com> # RK3588 Orange Pi 5 Plus, dual HDMI
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
---
 drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c | 33 +++++++++++++++++++++-----
 1 file changed, 27 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c b/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
index ecf12403101d..1c547016ee00 100644
--- a/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
+++ b/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
@@ -99,6 +99,7 @@ struct rockchip_hdmi_qp {
 	struct phy *phy;
 	struct gpio_desc *frl_enable_gpio;
 	struct delayed_work hpd_work;
+	int hpd_irq;
 	int port_id;
 	const struct rockchip_hdmi_qp_ctrl_ops *ctrl_ops;
 	unsigned long long tmds_char_rate;
@@ -308,13 +309,23 @@ static void dw_hdmi_qp_rk3588_enable_hpd(struct dw_hdmi_qp *dw_hdmi, void *data)
 	struct rockchip_hdmi_qp *hdmi = (struct rockchip_hdmi_qp *)data;
 
 	dw_hdmi_qp_rk3588_unmask_hpd(hdmi);
+	enable_irq(hdmi->hpd_irq);
 }
 
 static void dw_hdmi_qp_rk3588_disable_hpd(struct dw_hdmi_qp *dw_hdmi, void *data)
 {
 	struct rockchip_hdmi_qp *hdmi = (struct rockchip_hdmi_qp *)data;
 
+	disable_irq(hdmi->hpd_irq);
 	dw_hdmi_qp_rk3588_mask_hpd(hdmi);
+
+	/*
+	 * Cannot use the _sync() variant: the work takes mode_config.mutex,
+	 * which the connector probe path already holds when acquiring the
+	 * hpd_state_mutex this caller runs under. Requeuing is already ruled
+	 * out by disable_irq().
+	 */
+	cancel_delayed_work(&hdmi->hpd_work);
 }
 
 static const struct dw_hdmi_qp_phy_ops rk3588_hdmi_phy_ops = {
@@ -355,13 +366,23 @@ static void dw_hdmi_qp_rk3576_enable_hpd(struct dw_hdmi_qp *dw_hdmi, void *data)
 	struct rockchip_hdmi_qp *hdmi = (struct rockchip_hdmi_qp *)data;
 
 	dw_hdmi_qp_rk3576_unmask_hpd(hdmi);
+	enable_irq(hdmi->hpd_irq);
 }
 
 static void dw_hdmi_qp_rk3576_disable_hpd(struct dw_hdmi_qp *dw_hdmi, void *data)
 {
 	struct rockchip_hdmi_qp *hdmi = (struct rockchip_hdmi_qp *)data;
 
+	disable_irq(hdmi->hpd_irq);
 	dw_hdmi_qp_rk3576_mask_hpd(hdmi);
+
+	/*
+	 * Cannot use the _sync() variant: the work takes mode_config.mutex,
+	 * which the connector probe path already holds when acquiring the
+	 * hpd_state_mutex this caller runs under. Requeuing is already ruled
+	 * out by disable_irq().
+	 */
+	cancel_delayed_work(&hdmi->hpd_work);
 }
 
 static const struct dw_hdmi_qp_phy_ops rk3576_hdmi_phy_ops = {
@@ -584,7 +605,7 @@ static int dw_hdmi_qp_rockchip_bind(struct device *dev, struct device *master,
 	struct resource *res;
 	struct clk_bulk_data *clks;
 	struct clk *ref_clk;
-	int ret, irq, i;
+	int ret, i;
 
 	if (!dev->of_node)
 		return -ENODEV;
@@ -688,14 +709,14 @@ static int dw_hdmi_qp_rockchip_bind(struct device *dev, struct device *master,
 	if (plat_data.cec_irq < 0)
 		return plat_data.cec_irq;
 
-	irq = platform_get_irq_byname(pdev, "hpd");
-	if (irq < 0)
-		return irq;
+	hdmi->hpd_irq = platform_get_irq_byname(pdev, "hpd");
+	if (hdmi->hpd_irq < 0)
+		return hdmi->hpd_irq;
 
-	ret = devm_request_threaded_irq(dev, irq,
+	ret = devm_request_threaded_irq(dev, hdmi->hpd_irq,
 					cfg->ctrl_ops->hardirq_callback,
 					cfg->ctrl_ops->irq_callback,
-					IRQF_SHARED, "dw-hdmi-qp-hpd",
+					IRQF_NO_AUTOEN, "dw-hdmi-qp-hpd",
 					hdmi);
 	if (ret)
 		return ret;

-- 
2.55.0

[PATCH v12 15/16] drm/rockchip: dw_hdmi_qp: Use dw_hdmi_qp_hpd_notify() for HPD reports

[Cristian Ciocaltea]

Notify only the affected connector instead of polling all of them, and
ensure .detect_ctx() runs on reconnect to recover SCDC state.

Tested-by: Maud Spierings <maud_spierings@hotmail.com>
Tested-by: Diederik de Haas <diederik@cknow-tech.com>  # NanoPC-T6 LTS, Rock 5B
Tested-by: Robin Räber <robin.raebi@gmail.com>
Tested-by: Igor Paunovic <royalnet026@gmail.com> # RK3588 Orange Pi 5 Plus, dual HDMI
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
---
 drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c | 12 ++----------
 1 file changed, 2 insertions(+), 10 deletions(-)

diff --git a/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c b/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
index 1c547016ee00..f2a1ea8dfefa 100644
--- a/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
+++ b/drivers/gpu/drm/rockchip/dw_hdmi_qp-rockchip.c
@@ -24,7 +24,6 @@
 #include <drm/drm_bridge_connector.h>
 #include <drm/drm_managed.h>
 #include <drm/drm_of.h>
-#include <drm/drm_probe_helper.h>
 
 #include "rockchip_drm_drv.h"
 
@@ -398,14 +397,8 @@ static void dw_hdmi_qp_rk3588_hpd_work(struct work_struct *work)
 	struct rockchip_hdmi_qp *hdmi = container_of(work,
 						     struct rockchip_hdmi_qp,
 						     hpd_work.work);
-	struct drm_device *drm = hdmi->encoder.encoder.dev;
-	bool changed;
 
-	if (drm) {
-		changed = drm_helper_hpd_irq_event(drm);
-		if (changed)
-			dev_dbg(hdmi->dev, "connector status changed\n");
-	}
+	dw_hdmi_qp_hpd_notify(hdmi->hdmi);
 }
 
 static irqreturn_t dw_hdmi_qp_rk3576_hardirq(int irq, void *dev_id)
@@ -792,8 +785,7 @@ static int __maybe_unused dw_hdmi_qp_rockchip_resume(struct device *dev)
 
 	dw_hdmi_qp_resume(dev, hdmi->hdmi);
 
-	if (hdmi->encoder.encoder.dev)
-		drm_helper_hpd_irq_event(hdmi->encoder.encoder.dev);
+	dw_hdmi_qp_hpd_notify(hdmi->hdmi);
 
 	return 0;
 }

-- 
2.55.0

[PATCH v12 16/16] drm/bridge: dw-hdmi-qp: Drop unused .setup_hpd() phy op

[Cristian Ciocaltea]

With the transition to .{enable|disable}_hpd() phy ops complete in both
the bridge and platform drivers, .setup_hpd() has no remaining callers.
Drop it.

Tested-by: Maud Spierings <maud_spierings@hotmail.com>
Tested-by: Diederik de Haas <diederik@cknow-tech.com>  # NanoPC-T6 LTS, Rock 5B
Tested-by: Robin Räber <robin.raebi@gmail.com>
Tested-by: Igor Paunovic <royalnet026@gmail.com> # RK3588 Orange Pi 5 Plus, dual HDMI
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
---
 drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c | 3 ---
 include/drm/bridge/dw_hdmi_qp.h              | 1 -
 2 files changed, 4 deletions(-)

diff --git a/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c b/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c
index 54181b052fe2..b268c9376087 100644
--- a/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c
+++ b/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c
@@ -1089,9 +1089,6 @@ static void dw_hdmi_qp_init_hw(struct dw_hdmi_qp *hdmi)
 	/* Clear DONE and ERROR interrupts */
 	dw_hdmi_qp_write(hdmi, I2CM_OP_DONE_CLEAR | I2CM_NACK_RCVD_CLEAR,
 			 MAINUNIT_1_INT_CLEAR);
-
-	if (hdmi->phy.ops->setup_hpd)
-		hdmi->phy.ops->setup_hpd(hdmi, hdmi->phy.data);
 }
 
 struct dw_hdmi_qp *dw_hdmi_qp_bind(struct platform_device *pdev,
diff --git a/include/drm/bridge/dw_hdmi_qp.h b/include/drm/bridge/dw_hdmi_qp.h
index b4fb1c578a5b..ad33f9cafdeb 100644
--- a/include/drm/bridge/dw_hdmi_qp.h
+++ b/include/drm/bridge/dw_hdmi_qp.h
@@ -16,7 +16,6 @@ struct dw_hdmi_qp_phy_ops {
 	int (*init)(struct dw_hdmi_qp *hdmi, void *data);
 	void (*disable)(struct dw_hdmi_qp *hdmi, void *data);
 	enum drm_connector_status (*read_hpd)(struct dw_hdmi_qp *hdmi, void *data);
-	void (*setup_hpd)(struct dw_hdmi_qp *hdmi, void *data);
 	void (*enable_hpd)(struct dw_hdmi_qp *hdmi, void *data);
 	void (*disable_hpd)(struct dw_hdmi_qp *hdmi, void *data);
 };

-- 
2.55.0

Re: [PATCH v12 01/16] drm/bridge: dw-hdmi-qp: Demote i2c bus error messages to debug level

[Diederik de Haas]

Hi Cristian,

On Mon Oct 5, 2026 at 2:38 AM CEST, Cristian Ciocaltea wrote:
> A NACK or timeout on the DDC bus is a normal condition when the sink is
> absent or unpowered.  Some displays keep HPD asserted and the EDID
> EEPROM readable while in standby, but leave the SCDC slave unpowered.
> This results in i2c-related error messages on every boot during
> scrambling setup, even though the condition is handled gracefully and
> the SCDC configuration is restored on the next hotplug event.
>
> Log the i2c bus level failures with dev_dbg_ratelimited() instead,
> matching the silent behaviour of the non-QP dw-hdmi i2c bus algorithm.

I can confirm that with this patch I no longer see warnings like this:

  dwhdmiqp-rockchip fde80000.hdmi: i2c read error
  dwhdmiqp-rockchip fde80000.hdmi: i2c write nack!

When I add ``dw_hdmi_qp.dyndbg`` to the kernel command line, I do see
them at debug level.
This is when I boot up my NanoPC-T6 LTS connected to my 4K TV while it
is turned OFF.

When I boot up my NanoPC-T6 LTS connected to my 4K TV while it is
turned ON, dmesg is rather clean wrt dw_hdmi_qp both with and without
``dw_hdmi_qp.dyndbg`` in my kernel command line.

So it acts exactly as intended :-) Thus feel free to include my:

Tested-by: Diederik de Haas <diederik@cknow-tech.com>  # NanoPC-T6 LTS

Thanks!

Cheers,
  Diederik

> The DDC users already report failures at the appropriate level, e.g. the
> SCDC helpers via drm_scdc_dbg().  This also makes the no-hpd special
> casing of the read path messages redundant, so drop it.
>
> Reported-by: Diederik de Haas <diederik@cknow-tech.com>
> Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
> ---
>  drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c | 16 ++++------------
>  1 file changed, 4 insertions(+), 12 deletions(-)
>
> diff --git a/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c b/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c
> index c9568aea98dc..9abcea92b720 100644
> --- a/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c
> +++ b/drivers/gpu/drm/bridge/synopsys/dw-hdmi-qp.c
> @@ -350,22 +350,14 @@ static int dw_hdmi_qp_i2c_read(struct dw_hdmi_qp *hdmi,
>  
>  		stat = wait_for_completion_timeout(&i2c->cmp, HZ / 10);
>  		if (!stat) {
> -			if (hdmi->no_hpd)
> -				dev_dbg_ratelimited(hdmi->dev,
> -						    "i2c read timed out\n");
> -			else
> -				dev_err(hdmi->dev, "i2c read timed out\n");
> +			dev_dbg_ratelimited(hdmi->dev, "i2c read timed out\n");
>  			dw_hdmi_qp_write(hdmi, 0x01, I2CM_CONTROL0);
>  			return -EAGAIN;
>  		}
>  
>  		/* Check for error condition on the bus */
>  		if (i2c->stat & I2CM_NACK_RCVD_IRQ) {
> -			if (hdmi->no_hpd)
> -				dev_dbg_ratelimited(hdmi->dev,
> -						    "i2c read error\n");
> -			else
> -				dev_err(hdmi->dev, "i2c read error\n");
> +			dev_dbg_ratelimited(hdmi->dev, "i2c read error\n");
>  			dw_hdmi_qp_write(hdmi, 0x01, I2CM_CONTROL0);
>  			return -EIO;
>  		}
> @@ -404,14 +396,14 @@ static int dw_hdmi_qp_i2c_write(struct dw_hdmi_qp *hdmi,
>  
>  		stat = wait_for_completion_timeout(&i2c->cmp, HZ / 10);
>  		if (!stat) {
> -			dev_err(hdmi->dev, "i2c write time out!\n");
> +			dev_dbg_ratelimited(hdmi->dev, "i2c write time out!\n");
>  			dw_hdmi_qp_write(hdmi, 0x01, I2CM_CONTROL0);
>  			return -EAGAIN;
>  		}
>  
>  		/* Check for error condition on the bus */
>  		if (i2c->stat & I2CM_NACK_RCVD_IRQ) {
> -			dev_err(hdmi->dev, "i2c write nack!\n");
> +			dev_dbg_ratelimited(hdmi->dev, "i2c write nack!\n");
>  			dw_hdmi_qp_write(hdmi, 0x01, I2CM_CONTROL0);
>  			return -EIO;
>  		}