On MST links - at least for some MST branch devices - the list of modes
returned to users on an enabled link depends on the current link
rate/lane count parameters (besides the DPRX link capabilities, any MST
branch BW limit and the maximum link parameters reduced after LT
failures). In particular the MST branch BW limit may depend on the link
rate/lane count parameters programmed to DPCD. After an LT failure and
limiting the maximum link parameters accordingly, users should see a
mode list reflecting these new limits. However with the current fallback
order this isn't ensured, as the new limit could allow for modes
requiring a higher link BW, but these modes will be filtered out due to
the enabled link's lower link BW.
Ensure that the mode list changes in a consistent way after a link
training failure and reducing the link parameters by changing the
fallback order on MST links to happen in BW order.
v2:
- s/INTEL_DP_MAX_SUPPORTED_LANE_COUNTS/INTEL_DP_MAX_SUPPORTED_LANE_CONFIGS
and s/num_common_lane_counts/num_common_lane_configs to make the
difference wrt. max lane counts clearer. (Suraj)
- Add a TODO comment to make the SST fallback logic work the same way as
MST. (Arun)
- Use sort_r()'s default swap function instead of a custom one.
Cc: Suraj Kandpal <suraj.kandpal@intel.com>
Cc: Arun R Murthy <arun.r.murthy@intel.com>
Reviewed-by: Suraj Kandpal <suraj.kandpal@intel.com>
Reviewed-by: Arun R Murthy <arun.r.murthy@intel.com>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240729144458.2763667-1-imre.deak@intel.com
Add connector debugfs entries to force the link rate/lane count to be
used by a link training afterwards. These settings will be clamped to
the supported, i.e. the source's and sink's common rate/lane count.
After forcing the link rate/lane count reset the link training
parameters and for a non-auto setting disable reducing the link
parameters via the fallback logic. The former one can be used after
testing link training failure scenarios - via debugfs entries added
later - to reset the reduced link parameters after the test.
v2:
- Add the entries from intel_dp_link_training.c (Jani)
- Rename the entries to i915_dp_set_link_rate/lane_count.
v3: (Ville)
- Rename the entries/struct fields to force_link_rate/lane_count.
- Lock connection_mutex only for the required intel_dp state.
Cc: Jani Nikula <jani.nikula@intel.com>
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/20240610164933.2947366-18-imre.deak@intel.com
Recheck the link state after a passing link training, with a 2 sec delay
to account for cases where the link goes bad following the link training
and the sink doesn't report this via an HPD IRQ.
The delayed work added here will be also used by a later patch after a
failed link training to try to retrain the link with unchanged link
params before reducing the link params.
v2: Don't flush an uninitialized delayed work (on HDMI-only DDI ports).
v3:
- Move the helpers to a new intel_encoder.c file, rename them
accordingly. (Ville)
- Add the work to intel_encoder instead of intel_digital_port.
- Call the encoder specific link check function via an encoder hook.
- Flush the link check work during encoder destroy from
intel_dp_encoder_flush_work().
- Flush the link check work during encoder suspend as well.
v4: Call intel_encoder_link_check_init() with a valid encoder pointer.
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/20240610181428.2955658-1-imre.deak@intel.com
The DSC HW state of DP connectors is read out during driver loading and
system resume in intel_modeset_update_connector_atomic_state(). This
function is called for all connectors though and so the state of DSI
connectors will also get updated incorrectly, triggering a WARN there
wrt. the DSC decompression AUX device.
Fix the above by moving the DSC state readout to a new DP connector
specific sync_state() hook. This is anyway the logical place to update
the connector object's state vs. the connector's atomic state.
Fixes: b2608c6b32 ("drm/i915/dp_mst: Enable MST DSC decompression for all streams")
Reported-and-tested-by: Drew Davenport <ddavenport@chromium.org>
Closes: https://lore.kernel.org/all/Zb0q8IDVXS0HxJyj@chromium.org
Reviewed-by: Ankit Nautiyal <ankit.k.nautiyal@intel.com>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240205132631.1588577-1-imre.deak@intel.com
On shared (Thunderbolt) links with DP tunnels, the modeset may need to
be retried on all connectors on the link due to a link BW limitation
arising only after the atomic check phase. To support this add a helper
function queuing a work to retry the modeset on a given port's connector
and at the same time any MST connector with streams through the same
port. A follow-up change enabling the DP tunnel Bandwidth Allocation
Mode will take this into use.
v2:
- Send the uevent only to enabled MST connectors. (Jouni)
Cc: Jouni Högander <jouni.hogander@intel.com>
Reviewed-by: Uma Shankar <uma.shankar@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/20240220211841.448846-5-imre.deak@intel.com
Add hooks to intel_digital_port to lock and unlock the port and add a
helper to check the connector's detect status while the port is locked
already. This simplifies checking the connector detect status in
intel_dp_aux_xfer() and intel_digital_port_connected() in the next two
patches aborting AUX transfers on all DP connectors (except eDP) and
filtering HPD glitches.
Link: https://patchwork.freedesktop.org/patch/msgid/20240104083008.2715733-11-imre.deak@intel.com
Reviewed-by: Jouni Högander <jouni.hogander@intel.com>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Atm the allocated MST PBN value is calculated from the TU size (number
of allocated MTP slots) as
PBN = TU * pbn_div
pbn_div being the link BW for each MTP slot. For DP 1.4 link rates this
worked, as pbn_div there is guraranteed to be an integer number, however
on UHBR this isn't the case. To get a PBN, TU pair where TU is a
properly rounded-up value covering all the BW corresponding to PBN,
calculate first PBN and from PBN the TU value.
Calculate PBN directly from the effective pixel data rate, instead of
calculating it indirectly from the corresponding TU and pbn_div values
(which are in turn derived from the pixel data rate and BW overhead).
Add a helper function to calculate the effective data rate, also adding
a note that callers of intel_dp_link_required() may also need to check
the effective data rate (vs. the data rate w/o the BW overhead).
While at it add a note to check if WA#14013163432 is applicable.
v2:
- Fix PBN calculation, deriving it from the effective data rate directly
instead of using the indirect TU and pbn_div values for this.
- Add a note about WA#14013163432. (Arun)
v3:
- Fix rounding up quotient while calculating remote_tu. (Ville)
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Arun R Murthy <arun.r.murthy@intel.com> (v1)
Signed-off-by: Imre Deak <imre.deak@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20231117150929.1767227-3-imre.deak@intel.com
The link M/N ratio is the data rate / link symbol clock rate, fix things
up accordingly. On DP 1.4 this ratio was correct as the link symbol clock
rate in that case matched the link data rate (in bytes/sec units, the
symbol size being 8 bits), however it wasn't correct for UHBR rates
where the symbol size is 32 bits.
Kudos to Arun noticing in Bspec the incorrect use of link data rate in
the ratio's N value.
Reviewed-by: Arun R Murthy <arun.r.murthy@intel.com>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20231116131841.1588781-7-imre.deak@intel.com
Currently we seem to be using wrong DPCD register for reading
compressed bpps, reading min/max input bpc instead of compressed bpp.
Fix that, so that we now apply min/max compressed bpp limitations we
get from DP Spec Table 2-157 DP v2.0 and/or correspondent DPCD
register DP_DSC_MAX_BITS_PER_PIXEL_LOW/HIGH.
This might also allow us to get rid of an ugly compressed bpp
recalculation, which we had to add to make some MST hubs usable.
v2: - Fix operator precedence
v3: - Added debug info about compressed bpps
v4: - Don't try to intersect Sink input bpp and compressed bpps.
v5: - Decrease step while looking for suitable compressed bpp to
accommodate.
v6: - Use helper for getting min and max compressed_bpp (Ankit)
v7: - Fix checkpatch warning (Ankit)
Signed-off-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
Signed-off-by: Ankit Nautiyal <ankit.k.nautiyal@intel.com>
Reviewed-by: Ankit Nautiyal <ankit.k.nautiyal@intel.com>
Reviewed-by: Arun R Murthy <arun.r.murthy@intel.com>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20231030155843.2251023-31-imre.deak@intel.com
At the moment modesetting a stream CRTC will fail if the stream's BW
along with the current BW of all the other streams on the same MST link
is above the total BW of the MST link. Make the BW sharing more dynamic
by trying to reduce the link bpp of one or more streams on the MST link
in this case.
When selecting a stream to reduce the BW for, take into account which
link segment in the MST topology ran out of BW and which streams go
through this link segment. For instance with A,B,C streams in the same
MST topology A and B may share the BW of a link segment downstream of a
branch device, stream C not downstream of the branch device, hence not
affecting this BW. If this link segment's BW runs out one or both of
stream A/B's BW will be reduced until their total BW is within limits.
While reducing the link bpp for a given stream DSC may need to be
enabled for it, which requires FEC on the whole MST link. Check for this
condition and recompute the state for all streams taking the FEC
overhead into account (on 8b/10b links).
v2:
- Rebase on s/min_bpp_pipes/min_bpp_reached_pipes/ change.
Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20231030155843.2251023-29-imre.deak@intel.com
Enable DSC using the DSC AUX device stored for this purpose in the
connector. For clarity add separate functions to enable/disable the
decompression, since these sequences will diverge more in follow-up
patches that also enable/disable DSC passthrough and on MST do the
actual enabling/disabling only for the first/last user of the given
AUX device. As a preparation for the latter refcounting change, also
pass the atomic state to the functions.
While at it set/clear only the DP_DECOMPRESSION_EN flag in the
DP_DSC_ENABLE DPCD register, preserving the reserved register bits.
Besides preserving the reserved register bits, the behavior stays as
before, as DSC is still only enabled for the first MST stream (which a
follow-up patch changes, enabling it for all streams).
v2:
- Add a helper function setting/clearing the decompression flag,
preserving the reserved register bits.
v3:
- Add separate functions to enable/disable decompression and pass the
atomic state to these.
- Add DocBook for both functions.
Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> (v2)
Signed-off-by: Imre Deak <imre.deak@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20231030155843.2251023-24-imre.deak@intel.com
Enable FEC in crtc_state, as soon as it's known it will be needed by
DSC. This fixes the calculation of BW allocation overhead, in case DSC
is enabled by falling back to it during the encoder compute config
phase (vs. enabling FEC due to DSC being enabled on other streams).
v2:
- Enable FEC only in intel_dp_mst_find_vcpi_slots_for_bpp(), since
only by that will crtc_state->port_clock be set, which in turn is
needed by intel_dp_is_uhbr().
Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> (v1)
Signed-off-by: Imre Deak <imre.deak@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20231030155843.2251023-11-imre.deak@intel.com
The previous patches converted all users of the DSC DPCD caps to look
these up from the connector, so remove the version stored in intel_dp.
A follow-up patchset will read out the MST connector specific
capabilities in intel_dp_add_mst_connector() ->
intel_dp_mst_read_decompression_port_dsc_caps().
v2:
- Rebased on intel_edp_get_dsc_sink_cap() addition in the patchset.
v3:
- Rebased on read-out fix for eDP in the patchset.
Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> (v1)
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/20231011171606.2540078-3-imre.deak@intel.com
In non-DSC mode the link bpp can be set in 2*3 bpp steps in the pipe bpp
range, while in DSC mode it can be set in 1/16 bpp steps to any value
up to the maximum pipe bpp. Update the limits accordingly in both modes
to prepare for a follow-up patch which may need to reduce the max link
bpp value and starts to check the link bpp limits in DSC mode as well.
While at it add more detail to the link limit debug print and print it
also for DSC mode.
v2:
- Add to_bpp_frac_dec() instead of open coding it. (Jani)
v3: (Ville)
- Add BPP_X16_FMT / BPP_X16_ARG.
- Add TODO: comment about initializing the DSC link bpp limits earlier.
Cc: Jani Nikula <jani.nikula@linux.intel.com>
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/20230921195159.2646027-5-imre.deak@intel.com
A follow-up patch will need to limit the output link bpp both in the
non-DSC and DSC configuration, so track the pipe and link bpp limits
separately in the link_config_limits struct.
Use .4 fixed point format for link bpp matching the 1/16 bpp granularity
in DSC mode and for now keep this limit matching the pipe bpp limit.
v2: (Jani)
- Add to_bpp_int(), to_bpp_x16() helpers instead of opencoding them.
- Rename link_config_limits::link.min/max_bpp to min/max_bpp_x16.
Cc: Jani Nikula <jani.nikula@linux.intel.com>
Reviewed-by: Luca Coelho <luciano.coelho@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/20230921195159.2646027-3-imre.deak@intel.com
Currently there are many places where we use output_bpp for link bpp and
compressed bpp.
Lets use consistent naming:
output_bpp : The intermediate value taking into account the
output_format chroma subsampling.
compressed_bpp : target bpp for the DSC encoder.
link_bpp : final bpp used in the link.
For 444 sampling without DSC:
link_bpp = output_bpp = pipe_bpp
For 420 sampling without DSC:
output_bpp = pipe_bpp / 2
link_bpp = output_bpp
For 444 sampling with DSC:
output_bpp = pipe_bpp
link_bpp = compressed_bpp, computed with output_bpp (i.e. pipe_bpp in
this case)
For 420 sampling with DSC:
output_bpp = pipe_bpp/2
link_bpp = compressed_bpp, computed with output_bpp
Signed-off-by: Ankit Nautiyal <ankit.k.nautiyal@intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230817142459.89764-5-ankit.k.nautiyal@intel.com
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
Fix intel_dp_dsc_compute_config, previously timeslots parameter
was used in fact not as a timeslots, but more like a ratio
timeslots/64, which of course didn't have any effect for SST DSC,
but causes now issues for MST DSC.
Secondly we need to calculate pipe_bpp using intel_dp_dsc_compute_bpp
only for SST DSC case, while for MST case it has been calculated
earlier already with intel_dp_dsc_mst_compute_link_config.
Third we also were wrongly determining sink min bpp/max bpp, those
limites should be intersected with our limits to find common
acceptable bpp's, plus on top of that we should align those with
VESA bpps and only then calculate required timeslots amount.
Some MST hubs started to work only after third change was made.
v2: Make kernel test robot happy(claimed there was unitialzed use,
while there is none)
v3: Rename intel_dp_dsc_nearest_vesa_bpp to intel_dp_dsc_nearest_valid_bpp
(Manasi Navare)
Reviewed-by: Manasi Navare <manasi.d.navare@intel.com>
Signed-off-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20221123100718.29130-1-stanislav.lisovskiy@intel.com
We might to use that function separately from intel_dp_dsc_compute_config
for DP DSC over MST case, because allocating bandwidth in that
case can be a bit more tricky. So in order to avoid code copy-pasta
lets extract this to separate function and reuse it for both SST
and MST cases.
v2: Removed multiple blank lines
v3: Rename intel_dp_dsc_nearest_vesa_bpp to intel_dp_dsc_nearest_valid_bpp
to reflect its meaning more properly.
(Manasi Navare)
Signed-off-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
Reviewed-by: Manasi Navare <manasi.d.navare@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20221123100551.29080-1-stanislav.lisovskiy@intel.com
Whenever we are not able to get enough timeslots
for required PBN, let's try to allocate those
using DSC, just same way as we do for SST.
v2: Removed intel_dp_mst_dsc_compute_config and refactored
intel_dp_dsc_compute_config to support timeslots as a
parameter(Ville Syrjälä)
v3: - Rebased
- Added a debug to see that we at least try reserving
VCPI slots using DSC, because currently its not visible
from the logs, thus making debugging more tricky.
- Moved timeslots to numerator, where it should be.
v4: - Call drm_dp_mst_atomic_check already during link
config computation, because we need to know already
by this moment if uncompressed amount of VCPI slots
needed can fit, otherwise we need to use DSC.
(thanks to Vinod Govindapillai for pointing this out)
v5: - Put pipe_config->bigjoiner_pipes back to original
condition in intel_dp_dsc_compute_config
(don't remember when I lost it)
v6: - Removed unnecessary drm_dp_mst_atomic_check as it is
now always called in a newly introduced
intel_dp_mst_find_vcpi_slots_for_bpp function
(Vinod Govindapillai)
Reviewed-by: Vinod Govindapillai <vinod.govindapillai@intel.com>
Signed-off-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20221101094222.22091-5-stanislav.lisovskiy@intel.com
This reverts commit e1a84ba850.
Part of a series where patches were modified while applying to resolve
conflicts, leading to further conflicts between drm-misc-next and
drm-intel-next, resulting in build failures in drm-tip. To be applied
again on a baseline with drm-misc-next and drm-intel-next in sync.
Acked-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
