Instead of sleeping for a fixed 1ms (roughly, depending on timer slack),
start with a small sleep and exponentially increase the sleep on each
cycle.
A good example of a beneficiary is the guc mmio communication channel.
Typically we expect (and so spin) for 10us for a quick response, but this
doesn't cover everything and so sometimes we fallback to the millisecond+
sleep. This incurs a significant delay in time-critical operations like
preemption (igt/gem_exec_latency), which can be improved significantly by
using a small sleep after the spin fails.
We've made this suggestion many times, but had little experimental data
to support adding the complexity.
v2: Bump the minimum usleep to 10us on advice of
Documentation/timers/timers-howto.txt (Tvrko)
v3: Specify min, max range for usleep intervals -- some code may
crucially depend upon and so want to specify the sleep pattern.
References: 1758b90e38 ("drm/i915: Use a hybrid scheme for fast register waits")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: John Harrison <John.C.Harrison@intel.com>
Cc: Michał Winiarski <michal.winiarski@intel.com>
Cc: Ville Syrjala <ville.syrjala@linux.intel.com>
Reviewed-by: Michał Winiarski <michal.winiarski@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171124130031.20761-2-chris@chris-wilson.co.uk
The watermarks it should calculate against are the old optimal watermarks.
The currently active crtc watermarks are pure fiction, and are invalid in
case of a nonblocking modeset, page flip enabling/disabling planes or any
other reason.
When the crtc is disabled or during a modeset the intermediate watermarks
don't need to be programmed separately, and could be directly assigned
to the optimal watermarks.
CXSR must always be disabled in the intermediate case for modesets,
else we get a WARN for vblank wait timeout.
Also rename crtc_state to new_crtc_state, to distinguish it from the old
state.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171115163157.14372-2-maarten.lankhorst@linux.intel.com
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
The watermarks it should calculate against are the old optimal watermarks.
The currently active crtc watermarks are pure fiction, and are invalid in
case of a nonblocking modeset, page flip enabling/disabling planes or any
other reason.
When the crtc is disabled or during a modeset the intermediate watermarks
don't need to be programmed separately, and could be directly assigned
to the optimal watermarks.
CXSR must always be disabled in the intermediate case for modesets, else
we get a WARN for vblank wait timeout.
Also rename crtc_state to new_crtc_state, to distinguish it from the old state.
Changes since v1:
- Use intel_atomic_get_old_crtc_state. (ville)
Changes since v2:
- Always unset cxsr during modeset.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171115163157.14372-1-maarten.lankhorst@linux.intel.com
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Now that we always execute a context switch upon module load, there is
no need to queue a delayed task for doing so. The purpose of the delayed
task is to enable GT powersaving, for which we need the HW state to be
valid (i.e. having loaded a context and initialised basic state). We
used to defer this operation as historically it was slow (due to slow
register polling, fixed with commit 1758b90e38 ("drm/i915: Use a hybrid
scheme for fast register waits")) but now we have a requirement to save
the default HW state.
v2: Load the kernel context (to provide the power context) upon resume.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171112112738.1463-3-chris@chris-wilson.co.uk
Smatch warns of
drivers/gpu/drm/i915/intel_pm.c:1161 g4x_compute_wm() warn: signedness bug returning '(-33554430)'
which is a result of it believing that wm may be INT_MAX following
g4x_tlb_miss_wa(). Just declaring g4x_tlb_miss_wa() as returning an
unsigned integer is not sufficient, we need to tell smatch that wm itself
is unsigned for it to not worry. So mark up the locals we expect to be
non-negative, and so silence smatch.
v2: Mark up vlv_compute_wm_level() as unsigned similarly.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> #v1
Link: https://patchwork.freedesktop.org/patch/msgid/20171107140338.13748-1-chris@chris-wilson.co.uk
On CNL, individual wake rate limit was added to each engine.
GT can only go to RC6 if both Render and Media engines are
individually qualified. So we need to set their individual
wake rate limit.
+-----------------+---------------+--------------+--------------+
| | GT RC6 | Render C6 | Media C6 |
+-----------------+---------------+--------------+--------------+
| Wake rate limit | 0xA09C[31:16] | 0xA09C[15:0] | 0xA0A0[15:0] |
+-----------------+---------------+--------------+--------------+
v2: - Tune Render and Media wake rate values according to some extra
info I got from HW engineers. Value can be tuned, but for now
these are the recommended values.
- Fix typos pointed by James.
Cc: Nathan Ciobanu <nathan.d.ciobanu@intel.com>
Cc: Wayne Boyer <wayne.boyer@intel.com>
Cc: Joe Konno <joe.konno@linux.intel.com>
Cc: David Weinehall <david.weinehall@linux.intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: James Ausmus <james.ausmus@intel.com>
Reviewed-by: David Weinehall <david.weinehall@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171023224612.27208-1-rodrigo.vivi@intel.com
According to Spec for SKL+: "Isochronous Priority Control.
If enabled, Display sends demoted requests once the transition
watermark is reached. If transition watermark is not enabled,
Display sends demoted requests when the display buffer is full."
The commit 'e57f1c02155f ("drm/i915/gen9+: Add has_ipc flag in
device info structure")' introduced that as gen9+ but missing many
SKL Skus.
I believe the reason for that is Spec also mentions workarounds for
SKL-ALL: "IPC (Isoch Priority Control) may cause underflows
WA: Do not enable IPC in register ARB_CTL2"
It seems lame to add the feature and forever disable it,
but it will avoid a mistake of enabling it when we are reorganizing
the feature definitions on i915_pci.c later.
It will also allow us to probably extend that workaround for
other platforms.
Cc: Mahesh Kumar <mahesh1.kumar@intel.com>
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Mahesh Kumar <mahesh1.kumar@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171003063652.17248-1-rodrigo.vivi@intel.com
This patch adds IPC support. This patch also enables IPC in all supported
platforms based on has_ipc flag.
IPC (Isochronous Priority Control) is the hardware feature, which
dynamically controls the memory read priority of Display.
When IPC is enabled, plane read requests are sent at high priority until
filling above the transition watermark, then the requests are sent at
lower priority until dropping below the level 0 watermark.
The lower priority requests allow other memory clients to have better
memory access. When IPC is disabled, all plane read requests are sent at
high priority.
Changes since V1:
- Remove commandline parameter to disable ipc
- Address Paulo's comments
Changes since V2:
- Address review comments
- Set ipc_enabled flag
Changes since V3:
- move ipc_enabled flag assignment inside intel_ipc_enable function
Changes since V4:
- Re-enable IPC after suspend/resume
Changes since V5:
- Enable IPC for all gen >=9 except SKL
Changes since V6:
- fix commit msg
- after resume program IPC based on SW state.
Changes since V7:
- Modify IPC support check based on HAS_IPC macro (suggested by Chris)
Signed-off-by: Mahesh Kumar <mahesh1.kumar@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
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/20170817134529.2839-8-mahesh1.kumar@intel.com
GEN > 9 require transition WM to be programmed if IPC is enabled.
This patch calculates & enable transition WM for supported platforms.
If transition WM is enabled, Plane read requests are sent at high
priority until filling above the transition watermark, then the
requests are sent at lower priority until dropping below the level-0 WM.
The lower priority requests allow other memory clients to have better
memory access.
transition minimum is the minimum amount needed for trans_wm to work to
ensure the demote does not happen before enough data has been read to
meet the level 0 watermark requirements.
transition amount is configurable value. Higher values will
tend to cause longer periods of high priority reads followed by longer
periods of lower priority reads. Tuning to lower values will tend to
cause shorter periods of high and lower priority reads.
Keeping transition amount to 10 in this patch, as suggested by HW team.
Changes since V1:
- Address review comments from Maarten
Signed-off-by: Mahesh Kumar <mahesh1.kumar@intel.com>
Acked-by: Maarten Lankhorst <maarten.lankhorst@linux.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/20170817134529.2839-4-mahesh1.kumar@intel.com
Skip compressing 1 segment at the end of the frame,
avoid a pixel count mismatch nuke event when last active
pixel and dummy pixel has same color for Odd Plane
Width / Height.
For both platforms Gemini Lake and Cannon Lake.
v2: Use function-like macro and also use mask to clean
to make sure bit 11 is 0. (Suggested by Paulo).
v3: Add Display WA notation and also apply for GLK.
Both Forgotten on v2.
Using "GLK_" prefix since GLK came before CNL.
v4: Forgot to "|=" when moving directly macro to masked
val. (Noticed by Paulo.)
v5: Rebased on top of 0a46ddd57c ("drm/i915/cnp: Wa 1181:
Fix Backlight issue")
Cc: Imre Deak <imre.deak@intel.com>
Cc: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170905193013.31710-1-rodrigo.vivi@intel.com
If we miss the current vblank because the gpu was busy, that may cause a
jitter as the frame rate temporarily drops. We try to limit the impact
of this by then boosting the GPU clock to deliver the frame as quickly
as possible. Originally done in commit 6ad790c0f5 ("drm/i915: Boost GPU
frequency if we detect outstanding pageflips") but was never forward
ported to atomic and finally dropped in commit fd3a40242e ("drm/i915:
Rip out legacy page_flip completion/irq handling").
One of the most typical use-cases for this is a mostly idle desktop.
Rendering one frame of the desktop's frontbuffer can easily be
accomplished by the GPU running at low frequency, but often exceeds
the time budget of the desktop compositor. The result is that animations
such as opening the menu, doing a fullscreen switch, or even just trying
to move a window around are slow and jerky. We need to respond within a
frame to give the best impression of a smooth UX, as a compromise we
instead respond if that first frame misses its goal. The result should
be a near-imperceivable initial delay and a smooth animation even
starting from idle. The cost, as ever, is that we spend more power than
is strictly necessary as we overestimate the required GPU frequency and
then try to ramp down.
This of course is reactionary, too little, too late; nevertheless it is
surprisingly effective.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=102199
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Daniel Vetter <daniel.vetter@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170817123706.6777-1-chris@chris-wilson.co.uk
Tested-by: Lyude Paul <lyude@redhat.com>
Reviewed-by: Radoslaw Szwichtenberg <radoslaw.szwichtenberg@intel.com>
This bit enables hardware that will change the approximation used for distances
calculations for AA wide lines so that they are rendered more accurately.
The default value for this bit leaves the legacy behavior. There is no good
reason to not enable the new approximation except if comparing to previous GEN
rendered images.
v2: Rebase
v3: Fix author.
Rebased by Rodrigo who also added a comment as suggested by Oscar.
Since it is surrounded by Workarounds let's just add a comment to
make clear it is not an Wa.
Cc: Oscar Mateo <oscar.mateo@intel.com>
Signed-off-by: Ben Widawsky <benjamin.widawsky@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Oscar Mateo <oscar.mateo@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170815231651.975-4-rodrigo.vivi@intel.com
