Chris Wilson [Mon, 30 Apr 2018 13:15:02 +0000 (14:15 +0100)]
drm/i915: Retire requests along rings
In the next patch, rings are the central timeline as requests may jump
between engines. Therefore in the future as we retire in order along the
engine timeline, we may retire out-of-order within a ring (as the ring now
occurs along multiple engines), leading to much hilarity in miscomputing
the position of ring->head.
As an added bonus, retiring along the ring reduces the penalty of having
one execlists client do cleanup for another (old legacy submission
shares a ring between all clients). The downside is that slow and
irregular (off the critical path) process of cleaning up stale requests
after userspace becomes a modicum less efficient.
In the long run, it will become apparent that the ordered
ring->request_list matches the ring->timeline, a fun challenge for the
future will be unifying the two lists to avoid duplication!
v2: We need both engine-order and ring-order processing to maintain our
knowledge of where individual rings have completed upto as well as
knowing what was last executing on any engine. And finally by decoupling
retiring the contexts on the engine and the timelines along the rings,
we do have to keep a reference to the context on each request
(previously it was guaranteed by the context being pinned).
v3: Not just a reference to the context, but we need to keep it pinned
as we manipulate the rings; i.e. we need a pin for both the manipulation
of the engine state during its retirements, and a separate pin for the
manipulation of the ring state.
Chris Wilson [Mon, 30 Apr 2018 13:15:00 +0000 (14:15 +0100)]
drm/i915: Stop tracking timeline->inflight_seqnos
In commit 9b6586ae9f6b ("drm/i915: Keep a global seqno per-engine"), we
moved from a global inflight counter to per-engine counters in the
hope that will be easy to run concurrently in future. However, with the
advent of the desire to move requests between engines, we do need a
global counter to preserve the semantics that no engine wraps in the
middle of a submit. (Although this semantic is now only required for gen7
semaphore support, which only supports greater-then comparisons!)
v2: Keep a global counter of all requests ever submitted and force the
reset when it wraps.
Chris Wilson [Sat, 28 Apr 2018 11:15:32 +0000 (12:15 +0100)]
drm/i915/lrc: Scrub the GPU state of the guilty hanging request
Previously, we just reset the ring register in the context image such
that we could skip over the broken batch and emit the closing
breadcrumb. However, on resume the context image and GPU state would be
reloaded, which may have been left in an inconsistent state by the
reset. The presumption was that at worst it would just cause another
reset and skip again until it recovered, however it seems just as likely
to cause an unrecoverable hang. Instead of risking loading an incomplete
context image, restore it back to the default state.
v2: Fix up off-by-one from including the ppHSWP in with the register
state.
v3: Use a ring local to compact a few lines.
v4: Beware setting the ring local before checking for a NULL request.
References: https://bugs.freedesktop.org/show_bug.cgi?id=105304 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Cc: Michał Winiarski <michal.winiarski@intel.com> Cc: Michel Thierry <michel.thierry@intel.com> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Reviewed-by: Michel Thierry <michel.thierry@intel.com> #v2 Link: https://patchwork.freedesktop.org/patch/msgid/20180428111532.15819-1-chris@chris-wilson.co.uk
drm/i915/icl: update ddb entry start/end mask during hw ddb readout
Gen11/ICL onward ddb entry start/end mask is increased from 10 bits to
11 bits. This patch make changes to use proper mask for ICL+ during
hardware ddb value readout.
Changes since V1:
- Use _MASK & _SHIFT macro (James)
Changes since V2:
- use kernel type u8 instead of uint8_t
Changes since V3:
- Rebase
drm/i915/icl: Enable 2nd DBuf slice only when needed
ICL has two slices of DBuf, each slice of size 1024 blocks.
We should not always enable slice-2. It should be enabled only if
display total required BW is > 12GBps OR more than 1 pipes are enabled.
Changes since V1:
- typecast total_data_rate to u64 before multiplication to solve any
possible overflow (Rodrigo)
- fix where skl_wm_get_hw_state was memsetting ddb, resulting
enabled_slices to become zero
- Fix the logic of calculating ddb_size
Changes since V2:
- If no-crtc is part of commit required_slices will have value "0",
don't try to disable DBuf slice.
Changes since V3:
- Create a generic helper to enable/disable slice
- don't return early if total_data_rate is 0, it may be cursor only
commit, or atomic modeset without any plane.
Changes since V4:
- Solve checkpatch warnings
- use kernel types u8/u64 instead of uint8_t/uint64_t
Changes since V5:
- Rebase
This patch adds support to start tracking status of DBUF slices.
This is foundation to introduce support for enabling/disabling second
DBUF slice dynamically for ICL.
Changes Since V1:
- use kernel type u8 over uint8_t
James Ausmus [Wed, 28 Mar 2018 21:57:56 +0000 (14:57 -0700)]
drm/i915/icl: Don't set pipe CSC/Gamma in PLANE_COLOR_CTL
These fields have been deprecated and moved in ICL+. Stop setting the
bits.
They have moved to GAMMA_MODE and CSC_MODE, respectively. This patch
is just to stop incorrectly setting bits in PLANE_COLOR_CTL while
we're waiting for the new replacement functionality to be done.
v2: Drop useless comment, and change !(GEN >= 11) to (GEN < 11). (Ville)
drm/i915/psr/cnl: Set y-coordinate as valid in SDP
This was my bad, spec says that the name of this bit is
'Y-coordinate valid' but the values for it is:
0: Include Y-coordinate valid eDP1.4a
1: Do not include Y-coordinate valid eDP 1.4
So not setting it.
IGT tests could be improved with sink status, knowing for sure that
hardware have activate or exit PSR.
v3:
Reading i915_edp_psr_status was causing PSR to exit but now with
'drm/i915/psr: Prevent PSR exit when a non-pipe related register is
written' it is fixed.
drm/i915/psr: Prevent PSR exit when a non-pipe related register is written
Any write in any display register was causing HW to exit PSR,
masking it to allow more power savings. Writes to pipe related
registers will still cause HW to exit PSR.
This is already masked for PSR2.
It also do not break the Display WA #0884, writes to CURSURFLIVE
are still causing hardware to exit PSR. This was tested in CNL machine
by triggering a write to CURSURFLIVE when a debugfs was read by user.
Bspec: 7721 and 8042
v4: Checked that it do not breaks WA #0884 and added this information
to the commit message.
Chris Wilson [Wed, 11 Apr 2018 12:03:46 +0000 (13:03 +0100)]
drm/i915/selftests: Wait for idle between idle resets as well
Even though we weren't injecting guilty requests to be reset, we could
still fall over the issue of resetting the same request too fast -- where
the GPU refuses to start again. (Although it is interesting to note that
reloading the driver is sufficient, suggesting that we could recover if
we delayed the setup after reset?) Continue to paper over the problem by
adding a small delay by waiting for the engine to idle between tests,
and ensure that the engines are idle before starting the idle tests.
v2: Replace single instance of 50 with a magic macro.
References: 028666793a02 ("drm/i915/selftests: Avoid repeatedly harming the same innocent context") Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Cc: Michał Winiarski <michal.winiarski@intel.com> Cc: Michel Thierry <michel.thierry@intel.com> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Reviewed-by: Michał Winiarski <michal.winiarski@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180411120346.27618-1-chris@chris-wilson.co.uk
Jani Nikula [Thu, 26 Apr 2018 08:25:30 +0000 (11:25 +0300)]
drm/i915/dp: fix compliance test adjustments
Abstract compliance test adjustments to a single function. Also make the
bpc adjustments affect the limits, actually forcing the bpc. Seems like
directly changing the pipe_bpp in the past could not have been
effective.
Jani Nikula [Thu, 26 Apr 2018 08:25:26 +0000 (11:25 +0300)]
drm/i915/dp: abstract dp link config computation from the rest
Abstract a new intel_dp_compute_link_config() from
intel_dp_compute_config(), with the parts related to link configuration,
i.e. bpp, link rate, and lane count selection. No functional changes.
Jani Nikula [Thu, 26 Apr 2018 08:25:25 +0000 (11:25 +0300)]
drm/i915/dp: move link_bw and rate_select debugging where used
We call intel_dp_compute_rate() in intel_dp_compute_config() only to be
able to debug log the link_bw and rate_select parameters; we don't use
the parameters here for anything else. We call intel_dp_compute_rate()
again during link training where we actually need and use the
parameters.
Move the debug logging of link_bw and rate_select to
intel_dp_link_training_clock_recovery(), and clean up the extra
intel_dp_compute_rate() call and extra clutter from the already
overcrowded intel_dp_compute_config().
Chris Wilson [Thu, 26 Apr 2018 10:32:19 +0000 (11:32 +0100)]
drm/i915: Compile out engine debug for release
The majority of the engine state dumping is too voluminous to be useful
outside of a controlled setup, though a few do accompany severe errors.
Keep the debug dumps next to the errors, but hide the others behind a CI
compile flag. This becomes more useful when adding more dumps to latency
sensitive paths.
We can convert engine stats from a spinlock to seqlock to ensure interrupt
processing is never even a tiny bit delayed by parallel readers.
There is a smidgen bit more cost on the write lock side, and an extremely
unlikely chance that readers will have to retry a few times in face of
heavy interrupt load. But it should be extremely unlikely given how
lightweight read side section is compared to the interrupt processing
side, and also compared to the rest of the code paths which can lead into
it. Furthermore, writer is the ones doing the real, latency sensitive
work, while readers are only informative.
Oscar Mateo [Tue, 24 Apr 2018 21:39:55 +0000 (14:39 -0700)]
drm/i915/icl: Correctly clear lost ctx-switch interrupts across reset for Gen11
Interrupt handling in Gen11 is quite different from previous platforms.
v2: Rebased (Michel)
v3: Rebased with wiggle
v4: Rebased, remove TODO warning correctly (Daniele)
v5: Rebased, made gen11_gtiir const while at it (Michel)
v6: Rebased
v7: Adapt to the style currently in upstream
Suggested-by: Michel Thierry <michel.thierry@intel.com> Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Michel Thierry <michel.thierry@intel.com> Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Reviewed-by: Michel Thierry <michel.thierry@intel.com> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Link: https://patchwork.freedesktop.org/patch/msgid/1524605995-22324-1-git-send-email-oscar.mateo@intel.com
Chris Wilson [Wed, 25 Apr 2018 14:23:34 +0000 (15:23 +0100)]
drm/i915: Remove obsolete min/max freq setters from debugfs
A more complete, and more importantly stable, interface for controlling
the RPS frequency range is available in sysfs, obsoleting the unstable
debugfs.
It's presence seems to trick people into using it, forgetting it is not
ABI.
Chris Wilson [Wed, 25 Apr 2018 12:37:18 +0000 (13:37 +0100)]
drm/i915: Use memset64() to align the ring with MI_NOOP
When filling the ring to align the emit pointer to the next cacheline,
use memset64() rather than open-coding it. As we know that we always
have an even number of dwords, we can replace the dword loop with the
qword equivalent.
Chris Wilson [Tue, 24 Apr 2018 14:29:45 +0000 (15:29 +0100)]
drm/i915/breadcrumbs: Keep the fake irq armed across reset
Instead of synchronously cancelling the timer and re-enabling it inside
the reset callbacks, keep the timer enabled and let it die on its next
wakeup if no longer required. This allows
intel_engine_reset_breadcrumbs() to be used from an atomic
(timer/softirq) context such as required for resetting an engine.
It also allows us to react better to the user poking around debugfs for
testing missed irqs.
v2: Tighten the order of del_timer_sync as the fake_irq timer
may trigger the hangcheck timer, and so we should cancel it first and
then cancel the hangcheck (Mika)
There is a potential execution path in which variable err is
returned without being properly initialized previously.
Fix this by initializing variable err to 0.
Addresses-Coverity-ID: 1468362 ("Uninitialized scalar variable") Fixes: f4ecfbfc32ed ("drm/i915: Check whitelist registers across resets") Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.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/20180424131545.GA4053@embeddedor.com
drm/i915/fbdev: Enable late fbdev initial configuration
If the initial fbdev configuration (intel_fbdev_initial_config()) runs
and there still no sink connected it will cause
drm_fb_helper_initial_config() to return 0 as no error happened (but
internally the return is -EAGAIN). Because no framebuffer was
allocated, when a sink is connected intel_fbdev_output_poll_changed()
will not execute drm_fb_helper_hotplug_event() that would trigger
another try to do the initial fbdev configuration.
So here allowing drm_fb_helper_hotplug_event() to be executed when there
is no framebuffer allocated and fbdev was not set up yet.
This issue also happens when a MST DP sink is connected since boot, as
the MST topology is discovered in parallel if
intel_fbdev_initial_config() is executed before the first sink MST is
discovered it will cause this same issue.
This is a follow-up patch of
https://patchwork.freedesktop.org/patch/196089/
Changes from v1:
- not creating a dump framebuffer anymore, instead just allowing
drm_fb_helper_hotplug_event() to execute when fbdev is not setup yet.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=104158
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=104425 Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: stable@vger.kernel.org # v4.15+ Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: José Roberto de Souza <jose.souza@intel.com> Tested-by: Paul Menzel <pmenzel@molgen.mpg.de> Tested-by: frederik <frederik.schwan@linux.com> # 4.15.17 Tested-by: Ian Pilcher <arequipeno@gmail.com> Acked-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180418234158.9388-1-jose.souza@intel.com
Chris Wilson [Tue, 24 Apr 2018 08:16:00 +0000 (09:16 +0100)]
drm/i915: Don't dump umpteen thousand requests
If we have more than a few, possibly several thousand request in the
queue, don't show the central portion, just the first few and the last
being executed and/or queued. The first few should be enough to help
identify a problem in execution, and most often comparing the first/last
in the queue is enough to identify problems in the scheduling.
We may need some fine tuning to set MAX_REQUESTS_TO_SHOW for common
debug scenarios, but for the moment if we can avoiding spending more
than a few seconds dumping the GPU state that will avoid a nasty
livelock (where hangcheck spends so long dumping the state, it fires
again and starts to dump the state again in parallel, ad infinitum).
v2: Remember to print last not the stale rq iter after the loop.
Chris Wilson [Tue, 24 Apr 2018 01:08:39 +0000 (02:08 +0100)]
drm/i915: Build request info on stack before printk
printk unhelpfully inserts a '\n' between consecutive calls, and since
our drm_printf wrapper may be emitting info a seq_file instead,
KERN_CONT is not an option. To work with any drm_printf destination, we
need to build up the output into a temporary buf on the stack and then
feed the complete line in a single call to printk.
Mika Kuoppala [Mon, 23 Apr 2018 11:37:53 +0000 (14:37 +0300)]
drm/i915: Use ktime on wait_for
We use jiffies to determine when wait expires. However
Imre did find out that jiffies can and will do a >1
increments on certain situations [1]. When this happens
in a wait_for loop, we return timeout errorneously
much earlier than what the real wallclock would say.
We can't afford our waits to timeout prematurely.
Discard jiffies and change to ktime to detect timeouts.
Reported-by: Imre Deak <imre.deak@intel.com>
References: https://lkml.org/lkml/2018/4/18/798 [1]
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105771 Cc: Imre Deak <imre.deak@intel.com> Cc: Chris Wilson <chris@chris-wilson.co.uk> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: <stable@vger.kernel.org> Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Link: https://patchwork.freedesktop.org/patch/msgid/20180423113754.28424-1-mika.kuoppala@linux.intel.com
drm/i915/gvt: Mark expected switch fall-through in handle_g2v_notification
In preparation to enabling -Wimplicit-fallthrough, mark switch cases
where we are expecting to fall through.
Addresses-Coverity-ID: 1466154 ("Missing break in switch") Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com> Signed-off-by: Zhi Wang <zhi.a.wang@intel.com>
drm/i915/gvt/scheduler: Remove unnecessary NULL checks in sr_oa_regs
The checks are misleading and not required [1].
[1] https://lkml.org/lkml/2018/3/19/1792
Addresses-Coverity-ID: 1466017 Cc: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Zhi Wang <zhi.a.wang@intel.com>
Zhao Yan [Wed, 4 Apr 2018 05:57:09 +0000 (13:57 +0800)]
drm/i915/gvt: scan non-privileged batch buffer for debug purpose
For perfomance purpose, scanning of non-privileged batch buffer is turned
off by default. But for debugging purpose, it can be turned on via debugfs.
After scanning, we submit the original non-privileged batch buffer into
hardware, so that the scanning is only a peeking window of guest submitted
commands and will not affect the execution results.
v4:
- refine debugfs print format&content (zhenyu wang)
- print engine id instread of engine name to prevent potential memory leak
in debugfs warning message. (zhenyu wang)
v3:
- change vgpu->scan_nonprivbb from type bool to u32, so it is able to
selectively turn on/off scanning of non-privileged batch buffer on engine
level. e.g.
if vgpu->scan_nonprivbb=3, then it will scan non-privileged batch buffer
on engine 0 and 1.
- in debugfs interface to set vgpu->scan_nonprivbb, print warning message
to warn user and explicitly tell state change in kernel log (zhenyu wang)
v2:
- rebase
- update comments for start_gma_offset (henry)
Signed-off-by: Zhao Yan <yan.y.zhao@intel.com> Reviewed-by: Zhenyu Wang <zhenyuw@linux.intel.com> Signed-off-by: Zhi Wang <zhi.a.wang@intel.com>
Zhipeng Gong [Wed, 4 Apr 2018 00:43:53 +0000 (08:43 +0800)]
drm/i915/gvt: Update time slice more frequently
When there is only one vGPU in GVT-g and it submits workloads
continuously, it will not be scheduled out, vgpu_update_timeslice
is not called and its sched_in_time is not updated in a long time,
which can be several seconds or longer.
Once GVT-g pauses to submit workload for this vGPU due to heavy
host CPU workload, this vGPU get scheduled out and
vgpu_update_timeslice is called, its left_ts will be subtract
by a big value from sched_out_time - sched_in_time.
When GVT-g is going to submit workload for this vGPU again,
it will not be scheduled in until gvt_balance_timeslice reaches
stage 0 and reset its left_ts, which introduces several
hunderand milliseconds latency.
This patch updates time slice in every ms to update sched_in_time
timely.
v2: revise commit message
v3: use more concise expr. (Zhenyu)
Signed-off-by: Zhipeng Gong <zhipeng.gong@intel.com> Cc: Zhenyu Wang <zhenyuw@linux.intel.com> Cc: Min He <min.he@intel.com> Signed-off-by: Zhi Wang <zhi.a.wang@intel.com>
Zhipeng Gong [Wed, 4 Apr 2018 00:43:52 +0000 (08:43 +0800)]
drm/i915/gvt: Use real time to do timer check
intel_gvt_schedule check timer through a counter and is supposed
to wake up to increase the counter every ms.
In a system with heavy workload, gvt_service_thread can not get
a chance to run right after wake up and will be delayed several
milliseconds. As a result, one hundred counter interval means
several hundred milliseconds in real time.
This patch use real time instead of counter to do timer check.
Signed-off-by: Zhipeng Gong <zhipeng.gong@intel.com> Cc: Zhenyu Wang <zhenyuw@linux.intel.com> Cc: Min He <min.he@intel.com> Signed-off-by: Zhi Wang <zhi.a.wang@intel.com>
Hans de Goede [Fri, 20 Apr 2018 09:59:33 +0000 (11:59 +0200)]
drm/i915: Do NOT skip the first 4k of stolen memory for pre-allocated buffers v2
Before this commit the WaSkipStolenMemoryFirstPage workaround code was
skipping the first 4k by passing 4096 as start of the address range passed
to drm_mm_init(). This means that calling drm_mm_reserve_node() to try and
reserve the firmware framebuffer so that we can inherit it would always
fail, as the firmware framebuffer starts at address 0.
Commit d43537610470 ("drm/i915: skip the first 4k of stolen memory on
everything >= gen8") says in its commit message: "This is confirmed to fix
Skylake screen flickering issues (probably caused by the fact that we
initialized a ring in the first page of stolen, but I didn't 100% confirm
this theory)."
Which suggests that it is safe to use the first page for a linear
framebuffer as the firmware is doing (see note below).
This commit always passes 0 as start to drm_mm_init() and works around
WaSkipStolenMemoryFirstPage in i915_gem_stolen_insert_node_in_range()
by insuring the start address passed by to drm_mm_insert_node_in_range()
is always 4k or more. All entry points to i915_gem_stolen.c go through
i915_gem_stolen_insert_node_in_range(), so that any newly allocated
objects such as ring-buffers will not be allocated in the first 4k.
The one exception is i915_gem_object_create_stolen_for_preallocated()
which directly calls drm_mm_reserve_node() which now will be able to
use the first 4k.
This fixes the i915 driver no longer being able to inherit the firmware
framebuffer on gen8+, which fixes the video output changing from the
vendor logo to a black screen as soon as the i915 driver is loaded
(on systems without fbcon).
Some notes about the mapping of the BIOS framebuffer:
v1 led to some discussion if the assumption of the intel_display.c code
that the firmware framebuffer is a linear mapping of the stolen memory
starting at offset 0 is still correct, because that would mean that the
GOP does not implement the WaSkipStolenMemoryFirstPage workaround.
To verify this the following code was added at the end of
i915_gem_object_create_stolen_for_preallocated() :
pr_err("first ggtt entry before bind: 0x%016llx\n",
readq(dev_priv->ggtt.gsm));
ret = i915_vma_bind(vma,
HAS_LLC(dev_priv) ? I915_CACHE_LLC : I915_CACHE_NONE,
PIN_UPDATE);
pr_err("i915_vma_bind ret %d\n", ret);
pr_err("first ggtt entry after bind: 0x%016llx\n",
readq(dev_priv->ggtt.gsm));
Which prints the mapping of the first page, then does a vma_bind() to
force update the mapping with our linear view of the framebuffer and
then prints the mapping of the first page again.
On an Asrock B150M Pro4S/D3 mainboard with i5-6500 CPU this prints:
[ 1.651141] first ggtt entry before bind: 0x0000000078c00001
[ 1.651151] i915_vma_bind ret 0
[ 1.651152] first ggtt entry after bind: 0x0000000078c00083
There are no visual changes with this patch (BIOS vendor logo still
stays in place when we inherit the BIOS framebuffer), so the vma_bind()
does not impact which memory is being scanned out.
The address of the first ggtt entry matches with the start of stolen
and the i915_vma_bind call only changes the first gtt entry's flags,
or-ing in _PAGE_RW (BIT(1)) and PPAT_CACHED (BIT(7)), which perfectly
matches what we would expect based on gen8_pte_encode()'s behavior.
So it seems that the GOP indeed does NOT implement the wa and the i915's
code assuming a linear mapping at the start of stolen for the BIOS fb
still holds true for gen8+.
I've also tested this on a Cherry Trail based device (a GPD Win)
with identical results (the flags are 0x1b after the vma_bind
on CHT, which matches with I915_CACHE_NONE).
Changed in v2: No code changes, extended the commit message with the
verification that the intel_display.c BIOS framebuffer mapping is still
correct.
Interrupts other than the one for AUX errors are required only for debug,
so unmask them via debugfs when the user requests debug.
User can make such a request with
echo 1 > <DEBUG_FS>/dri/0/i915_edp_psr_debug
There are no locks to serialize PSR debug enabling from
irq_postinstall() and debugfs for simplicity. As irq_postinstall() is
called only during module initialization/resume and IGT subtests
aren't expected to modify PSR debug at those times, we should be safe.
Ville Syrjälä [Tue, 3 Apr 2018 21:24:18 +0000 (14:24 -0700)]
drm/i915: Enable edp psr error interrupts on bdw+
Plug in the bdw+ irq handling for PSR interrupts. bdw+ supports psr on
any transcoder in theory, though the we don't currenty enable PSR except
on the EDP transcoder.
v2: From DK
* Rebased on drm-tip
v3: Switched author to Ville based on IRC discussion.
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com> Reviewed-by: Jose Roberto de Souza <jose.souza@intel.com> Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180403212420.25007-2-dhinakaran.pandiyan@intel.com
Imre Deak [Thu, 19 Apr 2018 15:51:09 +0000 (18:51 +0300)]
drm/i915: Enable display WA#1183 from its correct spot
The DMC FW specific part of display WA#1183 is supposed to be enabled
whenever enabling DC5 or DC6, so move it to the DC6 enable function
from the DC6 disable function.
I noticed this after Daniel's patch to remove the unused
skl_disable_dc6() function.
Fixes: 53421c2fe99c ("drm/i915: Apply Display WA #1183 on skl, kbl, and cfl") Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: <stable@vger.kernel.org> Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180419155109.29451-1-imre.deak@intel.com
Ville Syrjälä [Fri, 16 Mar 2018 18:36:25 +0000 (20:36 +0200)]
drm/i915: Protect PIPE_CONF_CHECK macros with do {} while(0)
Make the PIPE_CONF_CHECK macros a bit more robust by wrapping them
in do {} while(0). Avoids funky sirprises when you try put an 'else'
after a PIPE_CONF_CHECK invocation...
drm/i915/audio: set minimum CD clock to twice the BCLK
In GLK when the device boots with only 1366x768 panel without audio, HDA
codec doesn't come up. In this case, the CDCLK is less than twice the
BCLK. Even though audio isn't being enabled, having a too low CDCLK
leads to audio probe failing altogether.
Require CDCLK to be at least twice the BLCK regardless of audio. This is
a minimal fix to improve things. Unfortunately, this a) leads to too
high CDCLK being used when audio is not used, and b) is still not enough
to fix audio probe when no outputs are connected at probe time.
The proper fix would be to increase CDCLK dynamically from the audio
component hooks.
v2:
- Address comment (Jani)
- New design approach
v3: - Typo fix on top of v1
v4 by Jani: rewrite commit message, add comment in code
Cc: stable@vger.kernel.org Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Dhinakaran Pandiyan <dhinakaran.pandiyan@gmail.com> Cc: Wenkai Du <wenkai.du@intel.com> Reviewed-by: Wenkai Du <wenkai.du@intel.com> Tested-by: Wenkai Du <wenkai.du@intel.com> Acked-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=102937 Signed-off-by: Abhay Kumar <abhay.kumar@intel.com> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180418103707.14645-1-jani.nikula@intel.com
Chris Wilson [Wed, 18 Apr 2018 18:40:52 +0000 (19:40 +0100)]
drm/i915: Pack params to engine->schedule() into a struct
Today we only want to pass along the priority to engine->schedule(), but
in the future we want to have much more control over the various aspects
of the GPU during a context's execution, for example controlling the
frequency allowed. As we need an ever growing number of parameters for
scheduling, move those into a struct for convenience.
v2: Move the anonymous struct into its own function for legibility and
ye olde gcc.
Chris Wilson [Wed, 18 Apr 2018 18:40:51 +0000 (19:40 +0100)]
drm/i915: Rename priotree to sched
Having moved the priotree struct into i915_scheduler.h, identify it as
the scheduling element and rebrand into i915_sched. This becomes more
useful as we start attaching more information we require to propagate
through the scheduler.
v2: Use i915_sched_node for future distinctiveness
Chris Wilson [Wed, 18 Apr 2018 18:40:50 +0000 (19:40 +0100)]
drm/i915: Move the priotree struct to its own headers
Over time the priotree has grown from a sorted list to a more
complicated structure for propagating constraints along the dependency
chain to try and resolve priority inversion. Start to segregate this
information from the rest of the request/fence tracking.
On Geminilake, sometimes audio card is not getting
detected after reboot. This is a spurious issue happening on
Geminilake. HW codec and HD audio controller link was going
out of sync for which there was a fix in i915 driver but
was not getting invoked for GLK. Extending this fix to GLK as well.
Tested by Du,Wenkai on GLK board.
Bspec: 21829
v2: Instead of checking GEN9_BC, BXT and GLK macros, use IS_GEN9 macro (Jani N)
Chris Wilson [Sat, 14 Apr 2018 09:12:33 +0000 (10:12 +0100)]
drm/i915: Call i915_perf_fini() on init_hw error unwind
We have to cleanup after i915_perf_init(), even on the error path, as it
passes a pointer into the module to the sysfs core. If we fail to
unregister the sysctl table, we leave a dangling pointer which then may
explode anytime later.
Fixes: 9f9b2792b6d3 ("drm/i915/perf: reuse timestamp frequency from device info") Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Lionel Landwerlin <lionel.g.landwerlin@intel.com> Cc: Matthew Auld <matthew.auld@intel.com> Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> Reviewed-by: Michal Wajdeczko <michal.wajdeczko@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180414091233.32224-1-chris@chris-wilson.co.uk
Chris Wilson [Sat, 14 Apr 2018 12:27:54 +0000 (13:27 +0100)]
drm/i915: Check whitelist registers across resets
Add a selftest to ensure that we restore the whitelisted registers after
rewrite the registers everytime they might be scrubbed, e.g. module
load, reset and resume. For the other volatile workaround registers, we
export their presence via debugfs and check in igt/gem_workarounds.
However, we don't export the whitelist and rather than do so, let's test
them directly in the kernel.
The test we use is to read the registers back from the CS (this helps us
be sure that the registers will be valid for MI_LRI etc). In order to
generate the expected list, we split intel_whitelist_workarounds_emit
into two phases, the first to build the list and the second to apply.
Inside the test, we only build the list and then check that list against
the hw.
v2: Filter out pre-gen8 as they do not have RING_NONPRIV.
v3: Drop unused engine parameter, no plans to use it now or future.
It looks that GuC does not actively use GUC_CTL_DEVICE_INFO parameter
where we are passing GT type and Core family values.
Let's stop/remove setup of this parameter and remove related
definitions.
v2: (this time without squashed HAX)
- New title and description
- Remove also GUC_CORE_FAMILY_* definitions (Michel)
v3:
- The removed define GUC_CTL_DEVICE_INFO has been restored (Michel)
- Updated description (Sagar)
v4: rebase
Signed-off-by: Piotr Piórkowski <piotr.piorkowski@intel.com> Cc: Sagar Arun Kamble <sagar.a.kamble@intel.com> Cc: Michał Winiarski <michal.winiarski@intel.com> Cc: John A Spotswood <john.a.spotswood@intel.com> Cc: Michal Wajdeczko <michal.wajdeczko@intel.com> Cc: Chris Wilson <chris@chris-wilson.co.uk> Cc: Michel Thierry <michel.thierry@intel.com> Acked-by: Michel Thierry <michel.thierry@intel.com> Reviewed-by: Sagar Arun Kamble <sagar.a.kamble@intel.com> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Link: https://patchwork.freedesktop.org/patch/msgid/20180413085245.57206-1-piotr.piorkowski@intel.com
Mika Kuoppala [Thu, 12 Apr 2018 14:58:02 +0000 (17:58 +0300)]
drm/i915/cnl: Use mmio access to context status buffer
Evidence indicates that Cannonlake HWSP is not coherent
as it should. Revert to using mmio access for now.
Testcase: igt/gem_ctx_switch
References: https://bugs.freedesktop.org/show_bug.cgi?id=105888 Cc: Chris Wilson <chris@chris-wilson.co.uk> Cc: Rafael Antognolli <rafael.antognolli@intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com> Acked-by: Chris Wilson <chris@chris-wilson.co.uk> Link: https://patchwork.freedesktop.org/patch/msgid/20180412145802.23313-1-mika.kuoppala@linux.intel.com
Jani Nikula [Wed, 11 Apr 2018 13:15:18 +0000 (16:15 +0300)]
drm/i915/bios: filter out invalid DDC pins from VBT child devices
The VBT contains the DDC pin to use for specific ports. Alas, sometimes
the field appears to contain bogus data, and while we check for it later
on in intel_gmbus_get_adapter() we fail to check the returned NULL on
errors. Oops results.
The simplest approach seems to be to catch and ignore the bogus DDC pins
already at the VBT parsing phase, reverting to fixed per port default
pins. This doesn't guarantee display working, but at least it prevents
the oops. And we continue to be fuzzed by VBT.
One affected machine is Dell Latitude 5590 where a BIOS upgrade added
invalid DDC pins.
Oscar Mateo [Tue, 10 Apr 2018 16:12:47 +0000 (09:12 -0700)]
drm/i915: Split out functions for different kinds of workarounds
There are different kind of workarounds (those that modify registers that
live in the context image, those that modify global registers, those that
whitelist registers, etc...) and they have different requirements in terms
of where they are applied and how. Also, by splitting them apart, it should
be easier to decide where a new workaround should go.
v2:
- Add multiple MISSING_CASE
- Rebased
v3:
- Rename mmio_workarounds to gt_workarounds (Chris, Mika)
- Create empty placeholders for BDW and CHV GT WAs
- Rebased
v4: Rebased
v5:
- Rebased
- FORCE_TO_NONPRIV register exists since BDW, so make a path
for it to achieve universality, even if empty (Chris)
Oscar Mateo [Tue, 10 Apr 2018 16:12:46 +0000 (09:12 -0700)]
drm/i915: Move a bunch of workaround-related code to its own file
This has grown to be a sizable amount of code, so move it to
its own file before we try to refactor anything. For the moment,
we are leaving behind the WA BB code and the WAs that get applied
(incorrectly) in init_clock_gating, but we will deal with it later.
v2: Use intel_ prefix for code that deals with the hardware (Chris)
v3: Rebased
v4:
- Rebased
- New license header
v5:
- Rebased
- Added some organisational notes to the file (Chris)
v6: Include DOC section in the documentation build (Jani)
Chris Wilson [Wed, 11 Apr 2018 10:39:29 +0000 (11:39 +0100)]
drm/i915/execlists: Set queue priority from secondary port
We can refine our current execlists->queue_priority if we inspect
ELSP[1] rather than the head of the unsubmitted queue. Currently, we use
the unsubmitted queue and say that if a subsequent request is more
important than the current queue, we will rerun the submission tasklet
to evaluate the need for preemption. However, we only want to preempt if
we need to jump ahead of a currently executing request in ELSP. The
second reason for running the submission tasklet is amalgamate requests
into the active context on ELSP[0] to avoid a stall when ELSP[0] drains.
(Though repeatedly amalgamating requests into the active context and
triggering many lite-restore is off question gain, the goal really is to
put a context into ELSP[1] to cover the interrupt.) So if instead of
looking at the head of the queue, we look at the context in ELSP[1] we
can answer both of the questions more accurately -- we don't need to
rerun the submission tasklet unless our new request is important enough
to feed into, at least, ELSP[1].
v2: Add some comments from the discussion with Tvrtko.
v3: More commentary to cross-reference queue_request()
References: f6322eddaff7 ("drm/i915/preemption: Allow preemption between submission ports") Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Michał Winiarski <michal.winiarski@intel.com> Cc: Michel Thierry <michel.thierry@intel.com> Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180411103929.27374-1-chris@chris-wilson.co.uk
Jani Nikula [Tue, 10 Apr 2018 09:12:48 +0000 (12:12 +0300)]
drm/i915/bios: remove duplicated code
Apparently caused by a merge fail at some point. Due to the nature of
the duplicated block, the second one will have no effect, and there's no
need to backport.
drm/i915/pmu: Inspect runtime PM state more carefully while estimating RC6
While thinking about sporadic failures of perf_pmu/rc6-runtime-pm* tests
on some CI machines I have concluded that: a) the PMU readout of RC6 can
race against runtime PM transitions, and b) there are other reasons than
being runtime suspended which can cause intel_runtime_pm_get_if_in_use to
fail.
Therefore when estimating RC6 the code needs to assert we are indeed in
suspended state, and if not, the best we can do is return the last known
RC6 value.
Without this check we can calculate the estimated value based on un-
initialized or inappropriate internal state, which can result in over-
estimation, or in any case incorrect value being returned.
v2:
* Re-arrange the code a bit to avoid second unlock and return branch.
(Chris Wilson)
v3:
* Insert some strategic blank lines and improve commit msg.
(Chris Wilson)
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Fixes: 1fe699e30113 ("drm/i915/pmu: Fix sleep under atomic in RC6 readout")
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105010 Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Chris Wilson <chris@chris-wilson.co.uk> Cc: Imre Deak <imre.deak@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Link: https://patchwork.freedesktop.org/patch/msgid/20180410112704.24462-1-tvrtko.ursulin@linux.intel.com
Chris Wilson [Tue, 10 Apr 2018 11:14:17 +0000 (12:14 +0100)]
drm/i915/guc: Replace %phn with %ph
%phn is not a valid specifier, and the trailing 'n' is being eaten by
the format-specifier and defaulting to the ' ' separator. Avoid angering
smatch by using the unknown specifier, and use the default we expect.
drivers/gpu/drm/i915/intel_guc_ct.c:616 ctb_read() warn: '%ph' cannot be followed by 'n'
drivers/gpu/drm/i915/intel_guc_ct.c:616 ctb_read() warn: '%ph' cannot be followed by 'n'
drivers/gpu/drm/i915/intel_guc_ct.c:616 ctb_read() warn: '%ph' cannot be followed by 'n'
drivers/gpu/drm/i915/intel_guc_ct.c:669 ct_handle_response() warn: '%ph' cannot be followed by 'n'
drivers/gpu/drm/i915/intel_guc_ct.c:679 ct_handle_response() warn: '%ph' cannot be followed by 'n'
drivers/gpu/drm/i915/intel_guc_ct.c:693 ct_handle_response() warn: '%ph' cannot be followed by 'n'
drivers/gpu/drm/i915/intel_guc_ct.c:707 ct_handle_response() warn: '%ph' cannot be followed by 'n'
drivers/gpu/drm/i915/intel_guc_ct.c:727 ct_process_request() warn: '%ph' cannot be followed by 'n'
drivers/gpu/drm/i915/intel_guc_ct.c:803 ct_handle_request() warn: '%ph' cannot be followed by 'n'
Chris Wilson [Tue, 10 Apr 2018 13:33:54 +0000 (14:33 +0100)]
drm/i915: Don't fiddle with rps/rc6 across GPU reset
Resetting the GPU doesn't affect the RPS/RC6 state, so we can stop
forcibly reloading the registers.
Ville suggested this many moons ago, I said at that time that sanitizing
was no harm and meant that our bookkeeping was kept consistent with the
HW. However, in a forthcoming series, we want to split rps/rc6 GT
powermanagement and one of the key simplifications is the control of
when we enable it. Performing a crude sanitize in the middle of
i915_gem_reset() is then a huge wart.
drm/i915: Add debugfs file to clear FIFO underruns.
Adding a i915_fifo_underrun_reset debugfs file will make it possible
for IGT tests to clear FIFO underrun fallout at the start of each
subtest, and make re-enable FBC so tests always have maximum exposure
to features used by IGT. FIFO underruns and FBC bugs will no longer
hide when an earlier subtests disables both.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
References: https://bugs.freedesktop.org/show_bug.cgi?id=105685
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105681 Link: https://patchwork.freedesktop.org/patch/msgid/20180328100526.36467-1-maarten.lankhorst@linux.intel.com Acked-by: Jani Nikula <jani.nikula@linux.intel.com>
[mlankhorst: Reset FBC reason if underrun had occurred. (vivijim)] Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Chris Wilson [Thu, 5 Apr 2018 11:49:15 +0000 (12:49 +0100)]
drm/i915/psr: Chase psr.enabled only under the psr.lock
Inside the psr work function, we want to wait for PSR to idle first and
wish to do so without blocking the normal modeset path, so we do so
without holding the PSR lock. However, we first have to find which pipe
PSR was enabled on, which requires chasing into the PSR struct and
requires locking to prevent intel_psr_disable() from concurrently
setting our pointer to NULL.
Fixes: 995d30477496 ("drm/i915: VLV/CHV PSR Software timer mode") Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Durgadoss R <durgadoss.r@intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: <stable@vger.kernel.org> # v4.0+ Reviewed-by: Jose Roberto de Souza <jose.souza@intel.com> Acked-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180405114915.29609-1-chris@chris-wilson.co.uk
Imre Deak [Mon, 9 Apr 2018 12:27:16 +0000 (15:27 +0300)]
drm/i915/gen9_lp: Increase DDI PHY0 power well enabling timeout
On GLK sporadic timeouts occur during PHY0 enabling. Based on logs it looks
like they happen sometime after a system suspend/resume cycle, with the
same power well enabling succeeding both before and after the failed
one and no other problems observed. The current timeout in the code is
not actually specified by BSpec, so let's try to increase that until a
BSpec update.
This patch updates scaler max limit support for NV12
v2: Rebased (me)
v3: Rebased (me)
v4: Missed the Tested-by/Reviewed-by in the previous series
Adding the same to commit message in this version.
v5: Addressed review comments from Ville and rebased
- calculation of max_scale to be made
less convoluted by splitting it up a bit
- Indentation errors to be fixed in the series
v6: Rebased (me)
Fixed review comments from Paauwe, Bob J
Previous version, where a split of calculation
was done, was wrong. Fixed that issue here.
v7: Rebased (me)
v8: Rebased (me)
v9: Rebased (me)
v10: Rebased (me)
v11: Addressed review comments from Shashank Sharma
Alignment issues fixed.
When call to skl_update_scaler is made, 0 was being
sent instead of pixel_format.
When crtc update scaler is called, we dont have the
fb to derive the pixel format. Added the function
parameter bool plane_scaler_check to account for this.
v12: Fixed failure in IGT debugfs_test.
fb is NULL in skl_update_scaler_plane
Due to this, accessing fb->format caused failure.
Patch checks fb before using.
v13: In the previous version there was a flaw.
In skl_update_scaler during plane_scaler_check
if the format was non-NV12, it would set need_scaling
to false. This could reset the previously set need_scaling
from a previous condition check. Patch fixes this.
Patch also adds minimum src height for YUV 420 formats
to 16 (as defined in BSpec) and adds for checking this
range.
v14: Addressed review comments from Maarten
Just add a check for NV12 min src height in
skl_update_scaler and retain the remaining checks
as is. Added Reviewed By from Juha-Pekka Heikkila.
v15: Rebased the series.
v16: Changed fb height restriction to be >= 16 as per
Bspec. Earlier it was > 16.
v17: Adding src width and height to be mult of 4 restriction
to avoid pipe fifo underruns for NV12.
Credits-to: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Tested-by: Clinton Taylor <clinton.a.taylor@intel.com> Reviewed-by: Clinton Taylor <clinton.a.taylor@intel.com> Reviewed-by: Juha-Pekka Heikkila <juhapekka.heikkila@gmail.com> Signed-off-by: Chandra Konduru <chandra.konduru@intel.com> Signed-off-by: Nabendu Maiti <nabendu.bikash.maiti@intel.com> Signed-off-by: Vidya Srinivas <vidya.srinivas@intel.com> Reviewed-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/1523245273-30264-15-git-send-email-vidya.srinivas@intel.com
This patch adds NV12 to format_is_yuv() function
for sprite planes.
v2:
-Use intel_ prefix for format_is_yuv (Ville)
v3: Rebased (me)
v4: Rebased and addressed review comments from Clinton A Taylor.
"static function in intel_sprite.c is not available
to the primary plane functions".
Changed commit message - function modified for
sprite planes.
v5: Missed the Tested-by/Reviewed-by in the previous series
Adding the same to commit message in this version.
v6: Rebased (me)
v7: Rebased (me)
v8: Rebased (me)
v9: Rebased (me)
v10: Changed intel_format_is_yuv function from
static to non-static. We need to use it later from
other files for check.
v11: Rebased the patch. format_is_yuv has already
been renamed to intel_format_is_yuv in the color
patch series which is already merged. This function
which was previously static has already been made
non-static. So this patch after rebase just adds
NV12 to intel_format_is_yuv function.
v12: Added reviewed by from Juha-Pekka Heikkila
v13/v14/v15: Rebased the series
Tested-by: Clinton Taylor <clinton.a.taylor@intel.com> Reviewed-by: Juha-Pekka Heikkila <juhapekka.heikkila@gmail.com> Reviewed-by: Clinton Taylor <clinton.a.taylor@intel.com> Reviewed-by: Shashank Sharma <shashank.sharma@intel.com> Signed-off-by: Chandra Konduru <chandra.konduru@intel.com> Signed-off-by: Nabendu Maiti <nabendu.bikash.maiti@intel.com> Signed-off-by: Vidya Srinivas <vidya.srinivas@intel.com> Reviewed-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/1523245273-30264-14-git-send-email-vidya.srinivas@intel.com
This patch sets appropriate scaler mode for NV12 format.
In this mode, skylake scaler does either chroma-upsampling or
chroma-upsampling and resolution scaling
v2: Review comments from Ville addressed
NV12 case to be checked first for setting
the scaler
v3: Rebased (me)
v4: Rebased (me)
v5: Missed the Tested-by/Reviewed-by in the previous series
Adding the same to commit message in this version.
v6: Rebased (me)
v7: Rebased (me)
v8: Rebased (me)
Restricting the NV12 change for scaler to BXT and KBL
in this series.
v9: Rebased (me)
v10: As of now, NV12 has been tested on Gen9 and Gen10. However,
code is applicable to all GEN >= 9. Hence making
that change to keep it generic.
Comments under v8 is not valid anymore.
v11: Addressed review comments by Shashank Sharma.
For Gen10+, the scaler mode to be set it planar or normal
(single bit). Changed the code to be applicable to all
Gen.
v12: Addressed review comments from Shashank Sharma
For Gen9 (apart from GLK) bits 28:29 to be programmed
in PS_CTRL for NV12. For GLK and Gen10+, bit 29 to be set
for all Planar.
v13: Addressed review comments from Juha-Pekka Heikkila
"NV12 not to be supported by SKL"
Adding Reviewed by tag from Shashank Shamr
v14: Added reviewed by from Juha-Pekka Heikkila
v15: Rebased the series
Tested-by: Clinton Taylor <clinton.a.taylor@intel.com> Reviewed-by: Juha-Pekka Heikkila <juhapekka.heikkila@gmail.com> Reviewed-by: Shashank Sharma <shashank.sharma@intel.com> Reviewed-by: Clinton Taylor <clinton.a.taylor@intel.com> Signed-off-by: Chandra Konduru <chandra.konduru@intel.com> Signed-off-by: Nabendu Maiti <nabendu.bikash.maiti@intel.com> Signed-off-by: Vidya Srinivas <vidya.srinivas@intel.com> Reviewed-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/1523245273-30264-13-git-send-email-vidya.srinivas@intel.com
drm/i915: Enable YUV to RGB for Gen10 in Plane Ctrl Reg
If the fb format is YUV, enable the plane CSC mode bits
for the conversion.
v2: Addressed review comments from Shashank Sharma
Alignment issue fixed in i915_reg.h
v3: Adding Reviewed By from Shashank Sharma
v4: Rebased the patch. As part of rebasing, re-using
the color series defines which are already merged.
plane_state->base.color_encoding might not be set for
NV12. For now, just using PLANE_COLOR_CSC_MODE_YUV709_TO_RGB709
in glk_plane_color_ctl if format is NV12.
Display WA 827 applies to GEN9 (excluede GLK) and CNL.
Switching the plane format from NV12 to RGB and leaving system idle
results in display underrun and corruption.
WA: Set the bit 15 & bit 19 to 1b in the CLKGATE_DIS_PSL
register for the pipe in which NV12 plane is enabled.
v2: Addressed review comments from Maarten and
Juha-Pekka Heikkila. Added reviewed by from
Juha-Pekka Heikkila.
This patch splits skl_compute_wm/ddb functions into two parts.
One adds all affected pipes after the commit to atomic_state structure
and second part does compute the DDB.
Display Workaround #0826 (SKL:ALL BXT:ALL) & #1059(CNL:A)
Hardware sometimes fails to wake memory from pkg C states fetching the
last few lines of planar YUV 420 (NV12) planes. This causes
intermittent underflow and corruption.
WA: Disable package C states or do not enable latency levels 1 through 7
(WM1 - WM7) on NV12 planes.
drm/i915/skl+: make sure higher latency level has higher wm value
DDB allocation optimization algorithm requires/assumes ddb allocation for
any memory C-state level DDB value to be as high as level below the
current level. Render decompression requires level WM to be as high as
wm level-0. This patch fulfils both the requirements.
v2: Changed plane_num to plane_id in skl_compute_wm_levels
v3: Addressed review comments from Shashank Sharma
Changed the commit message "statement can be more clear,
"DDB value to be as high as level below " what is level below ?"
drm/i915/skl+: pass skl_wm_level struct to wm compute func
This patch passes skl_wm_level structure itself to watermark
computation function skl_compute_plane_wm function (instead
of its internal parameters). It reduces number of arguments
required to be passed.
NV12 requires WM calculation for UV plane as well.
UV plane WM should also fulfill all the WM related restrictions.
v2: Addressed review comments from Shashank Sharma.
v3: Addressed review comments from Shashank Sharma
Changed plane_num to plane_id in skl_compute_plane_wm_params
and skl_compute_plane_wm.
Adding reviewed by tag from Shashank Sharma
drm/i915/skl+: support verification of DDB HW state for NV12
For YUV 420 Planar formats like NV12,
buffer allocation is done for Y and UV surfaces separately.
For NV12 plane formats, the UV buffer
allocation must be programmed in the Plane Buffer Config register
and the Y buffer allocation must be programmed in the
Plane NV12 Buffer Config register. Both register values
should be verified during verify_wm_state.
Current code calculates DDB for planar formats in such a way that we
store DDB of plane-0 in plane 1 & vice-versa.
In order to make this clean this patch refactors WM/DDB calculation for
NV12 planar formats.
v2: Addressed review comments by Maarten
v3: Rebased and addressed review comments by Maarten
v4: Fixed a compilation issue of string replacement is_nv12 to
is_planar
drm/i915/skl+: rename skl_wm_values struct to skl_ddb_values
skl_wm_values struct contains values of pipe/plane DDB only.
so rename it for better readability of code. Similarly
skl_copy_wm_for_pipe copies DDB values.
Lyude Paul [Sat, 7 Apr 2018 01:10:53 +0000 (21:10 -0400)]
drm/i915/dp: Send DPCD ON for MST before phy_up
When doing a modeset where the sink is transitioning from D3 to D0 , it
would sometimes be possible for the initial power_up_phy() to start
timing out. This would only be observed in the last action before the
sink went into D3 mode was intel_dp_sink_dpms(DRM_MODE_DPMS_OFF). We
originally thought this might be an issue with us accidentally shutting
off the aux block when putting the sink into D3, but since the DP spec
mandates that sinks must wake up within 1ms while we have 100ms to
respond to an ESI irq, this didn't really add up. Turns out that the
problem is more subtle then that:
It turns out that the timeout is from us not enabling DPMS on the MST
hub before actually trying to initiate sideband communications. This
would cause the first sideband communication (power_up_phy()), to start
timing out because the sink wasn't ready to respond. Afterwards, we
would call intel_dp_sink_dpms(DRM_MODE_DPMS_ON) in
intel_ddi_pre_enable_dp(), which would actually result in waking up the
sink so that sideband requests would work again.
Since DPMS is what lets us actually bring the hub up into a state where
sideband communications become functional again, we just need to make
sure to enable DPMS on the display before attempting to perform sideband
communications.
Changes since v1:
- Remove comment above if (!intel_dp->is_mst) - vsryjala
- Move intel_dp_sink_dpms() for MST into intel_dp_post_disable_mst() to
keep enable/disable paths symmetrical
- Improve commit message - dhnkrn
Changes since v2:
- Only send DPMS off when we're disabling the last sink, and only send
DPMS on when we're enabling the first sink - dhnkrn
Changes since v3:
- Check against is_mst, not intel_dp->is_mst - dhnkrn/vsyrjala
Signed-off-by: Lyude Paul <lyude@redhat.com> Reviewed-by: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Tested-by: Laura Abbott <labbott@redhat.com> Cc: stable@vger.kernel.org Fixes: ad260ab32a4d9 ("drm/i915/dp: Write to SET_POWER dpcd to enable MST hub.") Link: https://patchwork.freedesktop.org/patch/msgid/20180407011053.22437-1-lyude@redhat.com
Chris Wilson [Fri, 6 Apr 2018 22:03:54 +0000 (23:03 +0100)]
drm/i915: Pass the set of guilty engines to i915_reset()
Currently, we rely on inspecting the hangcheck state from within the
i915_reset() routines to determine which engines were guilty of the
hang. This is problematic for cases where we want to run
i915_handle_error() and call i915_reset() independently of hangcheck.
Instead of relying on the indirect parameter passing, turn it into an
explicit parameter providing the set of stalled engines which then are
treated as guilty until proven innocent.
While we are removing the implicit stalled parameter, also make the
reason into an explicit parameter to i915_reset(). We still need a
back-channel for i915_handle_error() to hand over the task to the locked
waiter, but let's keep that its own channel rather than incriminate
another.
This leaves stalled/seqno as being private to hangcheck, with no more
nefarious snooping by reset, be it whole-device or per-engine. \o/
The only real issue now is that this makes it crystal clear that we
don't actually do any testing of hangcheck per se in
drv_selftest/live_hangcheck, merely of resets!
Chris Wilson [Fri, 6 Apr 2018 22:03:53 +0000 (23:03 +0100)]
drm/i915: Treat i915_reset_engine() as guilty until proven innocent
If we are resetting just one engine, we know it has stalled. So we can
pass the stalled parameter directly to i915_gem_reset_engine(), which
alleviates the necessity to poke at the generic engine->hangcheck.stalled
magic variable, leaving that under control of hangcheck as its name
implies. Other than simplifying by removing the indirect parameter along
this path, this allows us to introduce new reset mechanisms that run
independently of hangcheck.
Chris Wilson [Fri, 6 Apr 2018 15:51:44 +0000 (16:51 +0100)]
drm/i915: Split out parking from the idle worker for reuse
We will want to park GEM before disengaging the drive^W^W^W unwedging.
Since we already do the work for idling, expose the guts as a new
function that we can then reuse.
v2: Just skip if already parked; makes it more forgiving to use by
future callers.
v3: Extract mark_busy, rename it to i915_gem_unpark and place it next to
i915_gem_park so that we can evaluate it for symmetry more easily.
Calling GEM from inside i915_request looks to be a bit of a layering
violation, for the moment I am imaging them as being notify_cb.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Michal Wajdeczko <michal.wajdeczko@intel.com> Cc: Sagar Arun Kamble <sagar.a.kamble@intel.com> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com> #v1 Reviewed-by: Michal Wajdeczko <michal.wajdeczko@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180406155144.27791-1-chris@chris-wilson.co.uk
Chris Wilson [Fri, 6 Apr 2018 10:09:50 +0000 (11:09 +0100)]
drm/i915/selftests: Rename wait_for_hang() to wait_until_running()
Tvrtko mentioned that wait_for_hang() was confusing as it does not
actually wait for the aforementioned hang, just until the request is
running and we are *ready* to inject a hang. A quick
s/wait_for_hang/wait_until_running/ removes that confusion without
having to rethink the naming scheme, immediately at least.
Suggested-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> 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: Michel Thierry <michel.thierry@intel.com> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180406100950.19033-1-chris@chris-wilson.co.uk