Chris Wilson [Tue, 31 Dec 2019 12:27:08 +0000 (12:27 +0000)]
drm/i915/gt: Restore coarse power gating
The coarse power gating was disabled as part of commit 2248a28384fe
("drm/i915/gen8+: Add RC6 CTX corruption WA") as a prelude to recover
from the context corruption; the power gating itself has no direct
impact on the RC6 context corruption. However, that recovery scheme was
never implemented due to difficult corner cases, and so we no longer need
to keep the power gating disabled.
Fixes: 2248a28384fe ("drm/i915/gen8+: Add RC6 CTX corruption WA") Closes: https://gitlab.freedesktop.org/drm/intel/issues/846 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Imre Deak <imre.deak@intel.com> Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Cc: Eero Tamminen <eero.t.tamminen@intel.com> Cc: Jon Bloomfield <jon.bloomfield@intel.com> Reviewed-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20191231122708.4025916-1-chris@chris-wilson.co.uk
Abdiel Janulgue [Tue, 31 Dec 2019 20:03:56 +0000 (20:03 +0000)]
drm/i915: Introduce remap_io_sg() to prefault discontiguous objects
Provide a way to set the PTE within apply_page_range for discontiguous
objects in addition to the existing method of just incrementing the pfn
for a page range.
Fixes: cc662126b413 ("drm/i915: Introduce DRM_I915_GEM_MMAP_OFFSET") Signed-off-by: Abdiel Janulgue <abdiel.janulgue@linux.intel.com> Cc: Chris Wilson <chris@chris-wilson.co.uk> Cc: Matthew Auld <matthew.auld@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Link: https://patchwork.freedesktop.org/patch/msgid/20191231200356.409475-1-chris@chris-wilson.co.uk
Matt Roper [Tue, 24 Dec 2019 23:15:21 +0000 (15:15 -0800)]
drm/i915/tgl: Assume future platforms will inherit TGL's SFC capability
Our usual i915 convention is to assume that future platforms will follow
the same behavior as the latest platform of today. The VDBOX/SFC
capabilities described here don't seem like something that should be
specific to TGL, so let's future-proof by making the test apply to all
gen12+ platforms.
Chris Wilson [Tue, 31 Dec 2019 12:08:57 +0000 (12:08 +0000)]
drm/i915/gt: Tweak flushes around ivb ppgtt
A small tweak to flush then invalidate appears to improve the
reliability of ppGTT switches on Ivybridge -- but does not improve
hsw/vlv bcs reliability.
Chris Wilson [Sun, 29 Dec 2019 18:31:47 +0000 (18:31 +0000)]
drm/i915/gt: Ensure that all new contexts clear STOP_RING
Set up the RING_MI_MODE in new contexts to clear the STOP_RING bit, just
in case they find it still set after a reset (as they are the first
contexts to be run).
Chris Wilson [Sat, 21 Dec 2019 14:49:17 +0000 (14:49 +0000)]
drn/i915: Break up long i915_buddy_free_list() with a cond_resched()
In the selftests, we may feed very long lists of blocks to be freed on
culmination of the tests. This coupled with kasan and other
malloc-tracing can make the kmem_cache_free() operation time consuming,
and doing many of those trigger soft lockup warnings. Break the list up
with a cond_resched().
Manasi Navare [Sat, 28 Dec 2019 03:12:04 +0000 (19:12 -0800)]
drm/i915/dp: Disable Port sync mode correctly on teardown
While clearing the Ports ync mode enable and master select bits
we need to clear the register completely instead of using disable masks
v3:
* Remove reg variable (Matt)
v2:
* Just write 0 to the reg (Ville)
* Rebase
Bugzilla: https://gitlab.freedesktop.org/drm/intel/issues/5 Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Jani Nikula <jani.nikula@intel.com> Fixes: 51528afe7c5e ("drm/i915/display/icl: Disable transcoder port sync as part of crtc_disable() sequence") Signed-off-by: Manasi Navare <manasi.d.navare@intel.com> Reviewed-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20191228031204.10189-3-manasi.d.navare@intel.com
Manasi Navare [Sat, 28 Dec 2019 03:12:03 +0000 (19:12 -0800)]
drm/i915/dp: Make port sync mode assignments only if all tiles present
Add an extra check before making master slave assignments for tiled
displays to make sure we make these assignments only if all tiled
connectors are present. If not then initialize the state to defaults
so it does a normal non tiled modeset without transcoder port sync.
v4:
deafulat port sync values in prepare_cleared_state (Ville)
v3:
* Default master trans to INVALID to avoid pipe mismatch
v2:
* Rename icl_add_sync_mode_crtcs
* Move this function just before .compute_config hook
* Check if DP before master slave assignments (Ville)
Manasi Navare [Sat, 28 Dec 2019 03:12:02 +0000 (19:12 -0800)]
drm/i915/dp: Make sure all tiled connectors get added to the state with full modeset
In case of tiled displays, all the tiles are linke dto each other
for transcoder port sync. So in intel_atomic_check() we need to make
sure that we add all the tiles to the modeset and if one of the
tiles needs a full modeset then mark all other tiles for a full modeset.
We also need to force modeset for all synced crtcs after fastset check.
v6:
* Add comments about why we do not call
drm_atomic_helper_check_modeset (Matt)
* Add FIXME for a corner case where tile info might vanish (Matt)
v5:
* Rebase
v4:
* Fix logic for modeset_synced_crtcs (Ville)
v3:
* Add tile checks only for Gen >11
v2:
* Change crtc_state scope, remove tile_grp_id (Ville)
* Use intel_connector_needs_modeset() (Ville)
* Add modeset_synced_crtcs (Ville)
* Make sure synced crtcs are forced full modeset
after fastset check (Ville)
Suggested-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: José Roberto de Souza <jose.souza@intel.com> Cc: Matt Roper <matthew.d.roper@intel.com>
Bugzilla: https://gitlab.freedesktop.org/drm/intel/issues/5 Signed-off-by: Manasi Navare <manasi.d.navare@intel.com> Reviewed-by: Matt Roper <matthew.d.roper@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20191228031204.10189-1-manasi.d.navare@intel.com
Manasi Navare [Wed, 11 Dec 2019 21:24:33 +0000 (13:24 -0800)]
drm/fbdev: Fallback to non tiled mode if all tiles not present
In case of tiled displays, if we hotplug just one connector,
fbcon currently just selects the preferred mode and if it is
tiled mode then that becomes a problem if rest of the tiles are
not present.
So in the fbdev driver on hotplug when we probe the client modeset,
if we dont find all the connectors for all tiles, then on a connector
with one tile, just fallback to the first available non tiled mode
to display over a single connector.
On the hotplug of the consecutive tiled connectors, if the tiled mode
no longer exists because of fbcon size limitation, then return
no modes for consecutive tiles but retain the non tiled mode
on the 0th tile.
Use the same logic in case of connected boot case as well.
This has been tested with Dell UP328K tiled monitor.
v3:
* Check Num tiled conns that are connected (Manasi)
v2:
* Set the modes on consecutive hotplugged tiles to no mode
if tiled mode is pruned (Dave)
v1:
* Just handle the 1st connector hotplug case
* v1 Reviewed-by: Dave Airlie <airlied@redhat.com>
Bugzilla: https://gitlab.freedesktop.org/drm/intel/issues/5 Suggested-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Suggested-by: Dave Airlie <airlied@redhat.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Dave Airlie <airlied@redhat.com> Signed-off-by: Manasi Navare <manasi.d.navare@intel.com> Reviewed-by: Dave Airlie <airlied@redhat.com> (v2) Link: https://patchwork.freedesktop.org/patch/msgid/20191211212433.18185-2-manasi.d.navare@intel.com
Manasi Navare [Wed, 11 Dec 2019 21:24:32 +0000 (13:24 -0800)]
drm: Handle connector tile support only for modes that match tile size
DRM Fb driver expects multiple CRTCs if it sees connector->has_tile
is set, but we need to handle tile support and look for multiple CRTCs
only for the modes that match the tile size. The other modes should
be able to be displayed without tile support or uisng single CRTC.
This patch adds the check to match the tile size with requested mode
to handle the tile support.
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Cc: Jani Nikula <jani.nikula@intel.com> Cc: Dave Airlie <airlied@redhat.com> Signed-off-by: Manasi Navare <manasi.d.navare@intel.com> Reviewed-by: Dave Airlie <airlied@redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20191211212433.18185-1-manasi.d.navare@intel.com
Lucas De Marchi [Tue, 24 Dec 2019 08:40:12 +0000 (00:40 -0800)]
drm/i915: prefer 3-letter acronym for tigerlake
We are currently using a mix of platform name and acronym to name the
functions. Let's prefer the acronym as it should be clear what platform
it's about and it's shorter, so it doesn't go over 80 columns in a few
cases. This converts tigerlake to tgl where appropriate.
Lucas De Marchi [Tue, 24 Dec 2019 08:40:11 +0000 (00:40 -0800)]
drm/i915: prefer 3-letter acronym for ivybridge
We are currently using a mix of platform name and acronym to name the
functions. Let's prefer the acronym as it should be clear what platform
it's about and it's shorter, so it doesn't go over 80 columns in a few
cases. This converts ivybridge to ivb where appropriate.
Lucas De Marchi [Tue, 24 Dec 2019 08:40:10 +0000 (00:40 -0800)]
drm/i915: prefer 3-letter acronym for broadwell
We are currently using a mix of platform name and acronym to name the
functions. Let's prefer the acronym as it should be clear what platform
it's about and it's shorter, so it doesn't go over 80 columns in a few
cases. This converts broadwell to bdw where appropriate.
Lucas De Marchi [Tue, 24 Dec 2019 08:40:09 +0000 (00:40 -0800)]
drm/i915: prefer 3-letter acronym for ironlake
We are currently using a mix of platform name and acronym to name the
functions. Let's prefer the acronym as it should be clear what platform
it's about and it's shorter, so it doesn't go over 80 columns in a few
cases. This converts ironlake to ilk where appropriate.
Lucas De Marchi [Tue, 24 Dec 2019 08:40:08 +0000 (00:40 -0800)]
drm/i915: prefer 3-letter acronym for icelake
We are currently using a mix of platform name and acronym to name the
functions. Let's prefer the acronym as it should be clear what platform
it's about and it's shorter, so it doesn't go over 80 columns in a few
cases. This converts icelake to icl where appropriate.
Lucas De Marchi [Tue, 24 Dec 2019 08:40:07 +0000 (00:40 -0800)]
drm/i915: prefer 3-letter acronym for cannonlake
We are currently using a mix of platform name and acronym to name the
functions. Let's prefer the acronym as it should be clear what platform
it's about and it's shorter, so it doesn't go over 80 columns in a few
cases. This converts cannonlake to cnl where appropriate.
Lucas De Marchi [Tue, 24 Dec 2019 08:40:06 +0000 (00:40 -0800)]
drm/i915: prefer 3-letter acronym for skylake
We are currently using a mix of platform name and acronym to name the
functions. Let's prefer the acronym as it should be clear what platform
it's about and it's shorter, so it doesn't go over 80 columns in a few
cases. This converts skylake to skl where appropriate.
Lucas De Marchi [Tue, 24 Dec 2019 08:40:05 +0000 (00:40 -0800)]
drm/i915: prefer 3-letter acronym for haswell
We are currently using a mix of platform name and acronym to name the
functions. Let's prefer the acronym as it should be clear what platform
it's about and it's shorter, so it doesn't go over 80 columns in a few
cases. This converts haswell to hsw where appropriate.
Lucas De Marchi [Tue, 24 Dec 2019 08:40:04 +0000 (00:40 -0800)]
drm/i915: prefer 3-letter acronym for pineview
We are currently using a mix of platform name and acronym to name the
functions. Let's prefer the acronym as it should be clear what platform
it's about and it's shorter, so it doesn't go over 80 columns in a few
cases. This converts pineview to pnv where appropriate.
v2: Add missing conversions in intel_pm.c (Matt Roper). While at it, fix
missing blank lines between structs that would otherwise trigger
checkpatch errors (Lucas)
Lukasz Fiedorowicz [Fri, 27 Dec 2019 13:37:48 +0000 (19:07 +0530)]
drm/i915/lmem: debugfs for LMEM details
Debugfs i915_gem_object is extended to enable the IGTs to
detect the LMEM's availability and the total size of LMEM.
v2: READ_ONCE is used [Chris]
v3: %pa is used for printing the resource [Chris]
v4: All regions' details added to debugfs [Chris]
v5: Macro for_each_mem_region added
name is initialized at region init [Chris]
Signed-off-by: Lukasz Fiedorowicz <lukasz.fiedorowicz@intel.com> Signed-off-by: Matthew Auld <matthew.auld@intel.com> Signed-off-by: Stuart Summers <stuart.summers@intel.com> Signed-off-by: Ramalingam C <ramalingam.c@intel.com> Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Link: https://patchwork.freedesktop.org/patch/msgid/20191227133748.4330-1-ramalingam.c@intel.com
Chris Wilson [Sat, 28 Dec 2019 11:12:55 +0000 (11:12 +0000)]
drm/i915: Restore very early GPU reset
In commit d03b224f4252 ("drm/i915/gt: Apply sanitiization just before
resume") the GT sanitization was pulled into the resume path as we need
to know the backend in order to do a full reset prior to resume.
However, it is still imperative that we scrub existing GPU state before
clobbering in our early setup, so restore a minimal GPU reset at the
start of our init sequence.
Matt Roper [Tue, 24 Dec 2019 01:20:25 +0000 (17:20 -0800)]
drm/i915: Add Wa_1408615072 and Wa_1407596294 to icl,ehl
Workaround database indicates we should disable clock gating of both the
vsunit and hsunit.
Bspec: 33450
Bspec: 33451 Cc: stable@kernel.vger.org Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Matt Atwood <matthew.s.atwood@intel.com> Cc: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20191224012026.3157766-3-matthew.d.roper@intel.com Reviewed-by: Lucas De Marchi <lucas.demarchi@intel.com>
Matt Roper [Tue, 24 Dec 2019 01:20:24 +0000 (17:20 -0800)]
drm/i915: Extend WaDisableDARBFClkGating to icl,ehl,tgl
WaDisableDARBFClkGating, now known as Wa_14010480278, has been added to
the workaround tables for ICL, EHL, and TGL so we need to extend our
platform test accordingly.
Bspec: 33450
Bspec: 33451
Bspec: 52890 Cc: stable@kernel.vger.org Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Matt Atwood <matthew.s.atwood@intel.com> Cc: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20191224012026.3157766-2-matthew.d.roper@intel.com Reviewed-by: Lucas De Marchi <lucas.demarchi@intel.com>
Chris Wilson [Fri, 27 Dec 2019 10:30:50 +0000 (10:30 +0000)]
drm/i915/selftests: Err out on coherency if initialisation failed
If gt initialisation failed, we are left with no engines to use for
coherency testing. Currently we bug out, which makes the actual error,
so fail more gracefully instead.
Chris Wilson [Wed, 25 Dec 2019 23:07:03 +0000 (23:07 +0000)]
drm/i915/gt: Stop poking at engine->serial at a high level
In record defaults, if we emit a request we expect a context switck on
parking. We need take no further action, so don't even mess with the low
level engine->serial where it is not warranted.
Chris Wilson [Mon, 23 Dec 2019 21:10:08 +0000 (21:10 +0000)]
drm/i915/gt: Flush other retirees inside intel_gt_retire_requests()
Our goal in wait_for_idle (intel_gt_retire_requests) is to the current
workload *and* their idle barriers. This requires us to notice the late
arrival of those, which is done by inspecting the list of active
timelines. However, if a concurrent retirer is running that new timeline
may not be added until after we drop the lock -- so flush concurrent
retirers before we take the lock and inspect the list.
José Roberto de Souza [Mon, 23 Dec 2019 01:06:54 +0000 (17:06 -0800)]
drm/i915/display: Add comment to a function that probably can be removed
This function is only called from port sync and it is identical to
what will be executed again in intel_update_crtc() over port sync
pipes.
If it is really necessary at least it deserves a better name and a
comment, leaving it to people working on port sync.
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Cc: Manasi Navare <manasi.d.navare@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: José Roberto de Souza <jose.souza@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20191223010654.67037-7-jose.souza@intel.com
José Roberto de Souza [Mon, 23 Dec 2019 01:06:52 +0000 (17:06 -0800)]
drm/i915/display: Prepare for fastset external dependencies check
MST and port sync have master and slaves pipes and it brings
dependencies between pipes to allow fastset.
For example if only MST master needs a modeset all of its slaves also
needs to do a modeset.
This patch adds the base for external dependencies check, the MST and
port sync bits will be added in another patches.
v3:
- moved handling to intel_atomic_check() this way is guarantee that
all pipes will have its state computed
v4:
- added a function to return if MST master neeeds modeset to simply
code in intel_atomic_check()
v5:
- fixed and moved code to check if MST master needs a modeset
v6:
- previons version of this patch was split into two patches
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Cc: Manasi Navare <manasi.d.navare@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: José Roberto de Souza <jose.souza@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20191223010654.67037-5-jose.souza@intel.com
José Roberto de Souza [Mon, 23 Dec 2019 01:06:51 +0000 (17:06 -0800)]
drm/i915/dp: Fix MST disable sequence
The disable sequence after wait for transcoder off was not correctly
implemented.
The MST disable sequence is basically the same for HSW, SKL, ICL and
TGL, with just minor changes for TGL.
With this last patch we finally fixed the hotplugs triggered by MST
sinks during the disable/enable sequence, those were causing source
to try to do a link training while it was not ready causing CPU pipe
FIFO underrrus on TGL.
v2: Only unsetting TGL_TRANS_DDI_PORT_MASK for TGL on the post
disable sequence
v4: Rebased, moved MST sequences to intel_mst_post_disable_dp()
BSpec: 4231
BSpec: 4163
BSpec: 22243
BSpec: 49190 Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: José Roberto de Souza <jose.souza@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20191223010654.67037-4-jose.souza@intel.com
José Roberto de Souza [Mon, 23 Dec 2019 01:06:50 +0000 (17:06 -0800)]
drm/i915/display: Always enables MST master pipe first
Due to DDB overlaps the pipe enabling sequence is not always crescent.
As the previous patch selects the smallest pipe/transcoder in the MST
stream to be master and it needs to be enabled first, this changes
were needed to guarantee that.
So first lets enable all pipes that do not need a fullmodeset as
those don't have any external dependency and those are the ones that
can overlap with each other.
Then on the second loop it will enable all the pipes that needs a
modeset and don't depends on other pipes like MST master
pipe/transcoder.
Then finally all the pipes that needs a modeset and have dependency
on other pipes, that at this point are alread enabled.
v3: rebased
v4:
- added check for modeset_pipes too to decide if is necessary for a
wait a vblank
- added DDB allocation overlap check for pipes that needs a modeset
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Manasi Navare <manasi.d.navare@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: José Roberto de Souza <jose.souza@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20191223010654.67037-3-jose.souza@intel.com
José Roberto de Souza [Mon, 23 Dec 2019 01:06:49 +0000 (17:06 -0800)]
drm/i915/tgl: Select master transcoder for MST stream
On TGL the blending of all the streams have moved from DDI to
transcoder, so now every transcoder working over the same MST port must
send its stream to a master transcoder and master will send to DDI
respecting the time slots.
So here adding all the CRTCs that shares the same MST stream if
needed and computing their state again, it will pick the lowest
pipe/transcoder among the ones in the same stream to be master.
Most of the time skl_commit_modeset_enables() enables pipes in a
crescent order but due DDB overlapping it might not happen, this
scenarios will be handled in the next patch.
v2:
- Using recently added intel_crtc_state_reset() to set
mst_master_transcoder to invalid transcoder for all non gen12 & MST
code paths
- Setting lowest pipe/transcoder as master, previously it was the
first one but setting a predictable one will help in future MST e
port sync integration
- Moving to intel type as much as we can
v3:
- Now intel_dp_mst_master_trans_compute() returns the MST master transcoder
- Replaced stdbool.h by linux/types.h
- Skip the connector being checked in
intel_dp_mst_atomic_master_trans_check()
- Using pipe instead of transcoder to compute MST master
v4:
- renamed connector_state to conn_state
v5:
- Improved the parameters of intel_dp_mst_master_trans_compute() to
simply code
- Added call drm_atomic_add_affected_planes() in
intel_dp_mst_atomic_master_trans_check() as helper could not do it
for us
- Removed "if (ret)" left over from v3 changes
v6:
- handled ret == I915_MAX_PIPES case in compute
intel_connector_needs_modeset() will be used outside of
intel_display.c in a future patch so it would only be necessary to
remove the state and add the prototype to the header file.
But while at it, I simplified the arguments and moved it to a better
place intel_atomic.c.
No behavior changes intended here.
v3:
- removed digital from exported version of intel_connector_needs_modeset
- rollback connector to drm type
v4:
- Renamed new_connector_state to new_conn_state
- Going back to drm_connector_state in
intel_encoders_update_prepare/complete as we also have
intel_tv_connector_state
Chris Wilson [Sun, 22 Dec 2019 23:35:58 +0000 (23:35 +0000)]
drm/i915: Mark the GEM context link as RCU protected
The only protection for intel_context.gem_cotext is granted by RCU, so
annotate it as a rcu protected pointer and carefully dereference it in
the few occasions we need to use it.
Chris Wilson [Sun, 22 Dec 2019 21:02:55 +0000 (21:02 +0000)]
drm/i915: Introduce a vma.kref
Start introducing a kref on i915_vma in order to protect the vma unbind
(i915_gem_object_unbind) from a parallel destruction (i915_vma_parked).
Later, we will use the refcount to manage all access and turn i915_vma
into a first class container.
Dhinakaran Pandiyan [Sat, 21 Dec 2019 12:05:41 +0000 (14:05 +0200)]
drm/i915: Skip rotated offset adjustment for unsupported modifiers
During framebuffer creation, we pre-compute offsets for 90/270 plane
rotation. However, only Y and Yf modifiers support 90/270 rotation. So,
skip the calculations for other modifiers.
To keep the gem buffer size check still working for tiled planes, factor
out the logic needed for rotation setup and skip only this part for
tiled planes other than Y/Yf.
v2: Add a bounds check WARN for the rotation info array.
v3: Keep the gem buffer size check working for tiled planes.
Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Mika Kahola <mika.kahola@intel.com> Signed-off-by: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Mika Kahola <mika.kahola@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20191221120543.22816-9-imre.deak@intel.com
Imre Deak [Sat, 21 Dec 2019 12:05:40 +0000 (14:05 +0200)]
drm/i915/tgl: Make sure FBs have a correct CCS plane stride
The CCS plane stride must be fixed on TGL, as it's not configurable for
the display. Instead the HW has a hardwired logic to determine it from
the main plane stride. Make sure userspace passes in the correct stride.
Dhinakaran Pandiyan [Sat, 21 Dec 2019 12:05:39 +0000 (14:05 +0200)]
drm/i915/tgl: Gen-12 render decompression
Gen-12 display decompression operates on Y-tiled compressed main surface.
The CCS is linear and has 4 bits of metadata for each main surface cache
line pair, a size ratio of 1:256. Gen-12 display decompression is
incompatible with buffers compressed by earlier GPUs, so make use of a new
modifier to identify gen-12 compression. Another notable change is that
render decompression is supported on all planes except cursor and on all
pipes. Start by adding render decompression support for [A,X]BGR888 pixel
formats.
v2: Fix checkpatch warnings (Lucas)
v3:
Rebase, disable color clear, styling changes and modify
intel_tile_width_bytes and intel_tile_height to handle linear CCS
v4:
- Use format block descriptors and the i915 specific func to get the
subsampling for each color plane.
- Use helpers to convert between CCS and main planes.
v5:
- Fix subsampling returned by intel_fb_plane_get_subsampling() for
the CCS plane of the first plane.
v6:
- Rebased on v2 of patch 4.
v7:
- Fix plane dimensions during FB check.
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Nanley G Chery <nanley.g.chery@intel.com> Cc: Jason Ekstrand <jason@jlekstrand.net> Cc: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Signed-off-by: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com> Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> (v6) Link: https://patchwork.freedesktop.org/patch/msgid/20191221120543.22816-7-imre.deak@intel.com
Dhinakaran Pandiyan [Sat, 21 Dec 2019 12:05:38 +0000 (14:05 +0200)]
drm/framebuffer: Format modifier for Intel Gen-12 render compression
Gen-12 has a new compression format, add a new modifier to indicate that.
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Nanley G Chery <nanley.g.chery@intel.com> Cc: Jason Ekstrand <jason@jlekstrand.net> Cc: Mika Kahola <mika.kahola@intel.com> Cc: dri-devel@lists.freedesktop.org Signed-off-by: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com> Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Mika Kahola <mika.kahola@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20191221120543.22816-6-imre.deak@intel.com
Imre Deak [Sat, 21 Dec 2019 12:05:37 +0000 (14:05 +0200)]
drm/i915: Add helpers to select correct ccs/aux planes
Using helpers instead of open coding this to select a CCS plane for a
main plane makes the code cleaner and less error-prone when the location
of CCS plane can be different based on the format (packed vs. YUV
semiplanar). The same applies to selecting an AUX plane which can be a
UV plane (for an uncompressed YUV semiplanar format), or a CCS plane.
Cc: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Mika Kahola <mika.kahola@intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Mika Kahola <mika.kahola@intel.com> Reviewed-by: Matt Roper <matthew.d.roper@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20191221120543.22816-5-imre.deak@intel.com
Easier to read if all the alignment changes are in one place and contained
within a function.
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Mika Kahola <mika.kahola@intel.com> Signed-off-by: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Mika Kahola <mika.kahola@intel.com> Reviewed-by: Matt Roper <matthew.d.roper@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20191221120543.22816-3-imre.deak@intel.com
Dhinakaran Pandiyan [Sat, 21 Dec 2019 12:05:34 +0000 (14:05 +0200)]
drm/i915: Use intel_tile_height() instead of re-implementing
intel_tile_dims() computes tile height using size and width, when there
is already a function to do just that - intel_tile_height()
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Mika Kahola <mika.kahola@intel.com> Signed-off-by: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Mika Kahola <mika.kahola@intel.com> Reviewed-by: Matt Roper <matthew.d.roper@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20191221120543.22816-2-imre.deak@intel.com
Chris Wilson [Sun, 22 Dec 2019 12:07:52 +0000 (12:07 +0000)]
drm/i915/gt: Pull GT initialisation under intel_gt_init()
Begin pulling the GT setup underneath a single GT umbrella; let intel_gt
take ownership of its engines! As hinted, the complication is the
lifetime of the probed engine versus the active lifetime of the GT
backends. We need to detect the engine layout early and keep it until
the end so that we can sanitize state on takeover and release.
Chris Wilson [Sat, 21 Dec 2019 20:01:09 +0000 (20:01 +0000)]
drm/i915: Move i915_gem_init_contexts() earlier
As the GEM global context setup is now independent of the GT state
(although GT does currently still depend upon the global
i915->kernel_context), we can move its init earlier, leaving the gt init
ready to be extracted.
Chris Wilson [Sat, 21 Dec 2019 18:02:04 +0000 (18:02 +0000)]
drm/i915/gt: Repeat wait_for_idle for retirement workers
Since we may retire timelines from secondary workers,
intel_gt_retire_requests() is not always a reliable indicator that all
pending retirements are complete. If we do detect secondary workers are
in progress, recommend intel_gt_wait_for_idle() to repeat the retirement
check.
Chris Wilson [Sat, 21 Dec 2019 16:03:24 +0000 (16:03 +0000)]
drm/i915: Remove i915->kernel_context
Allocate only an internal intel_context for the kernel_context, forgoing
a global GEM context for internal use as we only require a separate
address space (for our own protection).
Now having weaned GT from requiring ce->gem_context, we can stop
referencing it entirely. This also means we no longer have to create random
and unnecessary GEM contexts for internal use.
GEM contexts are now entirely for tracking GEM clients, and intel_context
the execution environment on the GPU.
Ville Syrjälä [Thu, 19 Dec 2019 11:14:30 +0000 (13:14 +0200)]
drm/i915: Introduce intel_crtc_state_alloc()
We have several places where we want to allocate a pristine
crtc state. Some of those currently call intel_crtc_state_reset()
to properly initialize all the non-zero defaults in the state, but
some places do not. Let's add intel_crtc_state_alloc() to do both
the alloc and the reset, and call that everywhere we need a fresh
crtc state.
v2: s/kzalloc/kmalloc/ since we memset() anyway (José)
Chris Wilson [Fri, 20 Dec 2019 10:12:30 +0000 (10:12 +0000)]
drm/i915: Push the use-semaphore marker onto the intel_context
Instead of rummaging through the intel_context to peek at the GEM
context in the middle of request submission to decide whether to use
semaphores, store that information on the intel_context itself.
Chris Wilson [Fri, 20 Dec 2019 10:12:29 +0000 (10:12 +0000)]
drm/i915: Drop GEM context as a direct link from i915_request
Keep the intel_context as being the primary state for i915_request, with
the GEM context a backpointer from the low level state for the rarer
cases we need client information. Our goal is to remove such references
to clients from the backend, and leave the HW submission agnostic to
client interfaces and self-contained.
Chris Wilson [Thu, 19 Dec 2019 23:29:32 +0000 (23:29 +0000)]
drm/i915/gt: Teach veng to defer the context allocation
Since we added the context_alloc callback to intel_context_ops, we can
safely install a custom hook for the deferred virtual context allocation.
This means that all new contexts behave the same upon creation,
simplifying later code.
Chris Wilson [Thu, 19 Dec 2019 22:13:44 +0000 (22:13 +0000)]
drm/i915/gt: Add breadcrumb retire to physical engine
Avoid adding the retire workers to the virtual engine so that we don't
end up in the unenviable situation of trying to free the virtual engine
while its worker remains active.
Fixes: dc93c9b69315 ("drm/i915/gt: Schedule request retirement when signaler idles") Closes: https://gitlab.freedesktop.org/drm/intel/issues/867 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Acked-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20191219221344.161523-1-chris@chris-wilson.co.uk
Ville Syrjälä [Fri, 13 Dec 2019 13:34:53 +0000 (15:34 +0200)]
drm/i915: Rename pipe update tracepoints
All the other display related tracepoints use intel_ instead
if i915_ as the prefix. Do the same for the pipe update
tracepoints so I don't always have to spend time looking for
them.
Ville Syrjälä [Fri, 13 Dec 2019 13:34:49 +0000 (15:34 +0200)]
drm/i915/fbc: Remove second redundant intel_fbc_pre_update() call
I fumbled the conflict resolution a bit when applying the
fbc vblank wait w/a. Because of that we now call intel_fbc_pre_update()
twice. Remove the second redundant call.
Colin Ian King [Thu, 19 Dec 2019 19:09:16 +0000 (19:09 +0000)]
drm/i915: fix uninitialized pointer reads on pointers to and from
Currently pointers to and from are not initialized and may contain
garbage values. This will cause uninitialized pointer reads in the
call to intel_frontbuffer_track and later checks to see if to and from
are null. Fix this by ensuring to and from are initialized to NULL.
Addresses-Coverity: ("Uninitialised pointer read)" Fixes: da42104f589d ("drm/i915: Hold reference to intel_frontbuffer as we track activity") Signed-off-by: Colin Ian King <colin.king@canonical.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Link: https://patchwork.freedesktop.org/patch/msgid/20191219190916.24693-1-colin.king@canonical.com
Chris Wilson [Wed, 18 Dec 2019 21:05:45 +0000 (21:05 +0000)]
drm/i915/gt: Suppress threshold updates on RPS parking
When we park RPS, we set the GPU to run at minimum 'idle' frequency.
However, as the GPU is idle, we also disable the worker and RPS
interrupts - changing the RPS thresholds has no effect, it just incurs
extra changes to restore them when we unpark. So on parking, leave the
thresholds set to the current power level and so we expect them to be
valid for our restart.
Chris Wilson [Wed, 18 Dec 2019 21:05:44 +0000 (21:05 +0000)]
drm/i915/gt: Use non-forcewake writes for RPS
Use non-forcewaked writes to queue RPS register changes that will take
effect when the write buffer is flushed, rather than wake the mmio
device for immediate effect. This is so that we can avoid a slow
forcewake dance upon unparking, and at our irregular updates.
Chris Wilson [Thu, 19 Dec 2019 12:43:53 +0000 (12:43 +0000)]
drm/i915/gt: Track engine round-trip times
Knowing the round trip time of an engine is useful for tracking the
health of the system as well as providing a metric for the baseline
responsiveness of the engine. We can use the latter metric for
automatically tuning our waits in selftests and when idling so we don't
confuse a slower system with a dead one.
Upon idling the engine, we send one last pulse to switch the context
away from precious user state to the volatile kernel context. We know
the engine is idle at this point, and the pulse is non-preemptible, so
this provides us with a good measurement of the round trip time. It also
provides us with faster engine parking for ringbuffer submission, which
is a welcome bonus (e.g. softer-rc6).
Chris Wilson [Thu, 19 Dec 2019 12:43:52 +0000 (12:43 +0000)]
drm/i915/gt: Schedule request retirement when signaler idles
Very similar to commit 4f88f8747fa4 ("drm/i915/gt: Schedule request
retirement when timeline idles"), but this time instead of coupling into
the execlists CS event interrupt, we couple into the breadcrumb
interrupt and queue a timeline's retirement when the last signaler is
completed. This should allow us to more rapidly park ringbuffer
submission, and so help reduce power consumption on older systems.
v2: Fixup intel_engine_add_retire() to handle concurrent callers
Jani Nikula [Thu, 12 Dec 2019 13:47:28 +0000 (15:47 +0200)]
drm/i915/dsc: fix DSC power domains for DSI
Fix several issues with DSC power domains that did not take DSI
transcoders into account:
- On TGL+ we need to use PW2 for DSC on pipe A, not transcoder A. There
is no longer an eDP transcoder, but there are two DSI transcoders
which may be connected to pipe A.
- On TGL+ we need to use the pipe, not transcoder, power domains for DSC
on pipes other than A. Again, there are DSI transcoders.
- On ICL we need to use PW2 for DSC also for DSI transcoders, not just
for the eDP transcoder.
Using is_pipe_dsc() also adds the warning about ICL pipe A DSC, which
does not exist.
Jani Nikula [Wed, 11 Dec 2019 16:23:47 +0000 (18:23 +0200)]
drm/i915/dsc: clarify DSC support for pipe A on ICL
The check for cpu_transcoder != TRANSCODER_A is more magic than
necessary, and potentially misleading. Before TGL, DSC is supported on
pipe A if, and only if, it's used with eDP or DSI transcoders. No
functional changes.
Chris Wilson [Wed, 18 Dec 2019 09:40:57 +0000 (09:40 +0000)]
drm/i915: Ratelimit i915_globals_park
When doing our global park, we like to be a good citizen and shrink our
slab caches (of which we have quite a few now), but each
kmem_cache_shrink() incurs a stop_machine() and so ends up being quite
expensive, causing machine-wide stalls. While ideally we would like to
throw away unused pages in our slab caches whenever it appears that we
are idling, doing so will require a much cheaper mechanism. In the
meantime use a delayed worked to impose a rate-limit that means we have
to have been idle for more than 2 seconds before we start shrinking.
Chris Wilson [Tue, 17 Dec 2019 09:56:41 +0000 (09:56 +0000)]
drm/i915/gt: Remove direct invocation of breadcrumb signaling
Only signal the breadcrumbs from inside the irq_work, simplifying our
interface and calling conventions. The micro-optimisation here is that
by always using the irq_work interface, we know we are always inside an
irq-off critical section for the breadcrumb signaling and can ellide
save/restore of the irq flags.
First is not assuming rc6 value zero means GT is asleep since that can
also mean GPU is fully busy and we do not want to enter the estimation
path in that case.
Second is ensuring monotonicity on the estimation path itself. I suspect
what is happening is with extremely rapid park/unpark cycles we get no
updates on the real rc6 and therefore have to careful not to
unconditionally trust use last known real rc6 when creating a new
estimation.
v2:
* Simplify logic by not tracking the estimate but last reported value.
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Fixes: 16ffe73c186b ("drm/i915/pmu: Use GT parked for estimating RC6 while asleep") Cc: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> # v1 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Link: https://patchwork.freedesktop.org/patch/msgid/20191217142057.1000-1-tvrtko.ursulin@linux.intel.com
Ville Syrjälä [Fri, 13 Dec 2019 19:52:17 +0000 (21:52 +0200)]
drm/i915: Move stuff from haswell_crtc_disable() into encoder .post_disable()
Move all of haswell_crtc_disable() into the encoder
.post_disable() hooks. Now we're left with just
calling the .disable() and .post_disable() hooks
back to back.
I chose to move the code into the .post_disable() hook instead
of the .disable() hook as most of the sequence is currently
implemented in the .post_disable() hook.
We should collapse it all down to just one hook and then the
encoders can drive the modeset sequence fully. But that may
need some further refactoring as we currently call the
ddi .post_disable() hook from mst code and we can't just
replace that with a call to the ddi .disable() hook.
Should also follow up with similar treatment for the enable
sequence but let's start here where it's easier.
Ville Syrjälä [Fri, 13 Dec 2019 19:52:16 +0000 (21:52 +0200)]
drm/i915: Pass old crtc state to intel_crtc_vblank_off()
To make life easier in the future let's pass the old crtc state
to intel_crtc_vblank_off() just like we already do for its
counterpart intel_crtc_vblank_on().
Ville Syrjälä [Fri, 13 Dec 2019 19:52:15 +0000 (21:52 +0200)]
drm/i915: Pass old crtc state to skylake_scaler_disable()
To make life easier in the future let's pass the old crtc state
to skylake_scaler_disable() just like we already do for
for its ancestor ironlake_pfit_disable().
projects / users / hch / configfs.git / log