drm/i915: Fix warnings while make xmldocs caused by intel_lrc.c
This patch fix following warnings while "make xmldocs".
.//drivers/gpu/drm/i915/intel_lrc.c:780: warning: No description
found for parameter 'req'
.//drivers/gpu/drm/i915/intel_lrc.c:780: warning: Excess function
parameter 'request' description in 'intel_logical_ring_begin'
.//drivers/gpu/drm/i915/intel_lrc.c:780: warning: Excess function
parameter 'ctx' description in 'intel_logical_ring_begin'
Signed-off-by: Masanari Iida <standby24x7@gmail.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Nick Hoath [Fri, 11 Sep 2015 11:53:46 +0000 (12:53 +0100)]
drm/i915: Split alloc from init for lrc
Extend init/init_hw split to context init.
- Move context initialisation in to i915_gem_init_hw
- Move one off initialisation for render ring to
i915_gem_validate_context
- Move default context initialisation to logical_ring_init
Rename intel_lr_context_deferred_create to
intel_lr_context_deferred_alloc, to reflect reduced functionality &
alloc/init split.
This patch is intended to split out the allocation of resources &
initialisation to allow easier reuse of code for resume/gpu reset.
v2: Removed function ptr wrapping of do_switch_context (Daniel Vetter)
Left ->init_context int intel_lr_context_deferred_alloc
(Daniel Vetter)
Remove unnecessary init flag & ring type test. (Daniel Vetter)
Improve commit message (Daniel Vetter)
v3: On init/reinit, set the hw next sequence number to the sw next
sequence number. This is set to 1 at driver load time. This prevents
the seqno being reset on reinit (Chris Wilson)
v4: Set seqno back to ~0 - 0x1000 at start-of-day, and increment by 0x100
on reset.
This makes it obvious which bbs are which after a reset. (David Gordon
& John Harrison)
Rebase.
v5: Rebase. Fixed rebase breakage. Put context pinning in separate
function. Removed code churn. (Thomas Daniel)
v6: Cleanup up issues introduced in v2 & v5 (Thomas Daniel)
Issue: VIZ-4798 Signed-off-by: Nick Hoath <nicholas.hoath@intel.com> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: Chris Wilson <chris@chris-wilson.co.uk> Cc: John Harrison <john.c.harrison@intel.com> Cc: David Gordon <david.s.gordon@intel.com> Cc: Thomas Daniel <thomas.daniel@intel.com> Reviewed-by: Thomas Daniel <thomas.daniel@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
drm/i915/bxt: Use intel_encoder->hpd_pin to check live status
Using intel_encoder's hpd_pin to check the live status
because of BXT A0/A1 WA for HPD pins and hpd_pin contains the
updated pin for the corresponding port.
Michel Thierry [Fri, 4 Sep 2015 11:59:15 +0000 (12:59 +0100)]
drm/i915/lrc: Prevent preemption when lite-restore is disabled
When WaEnableForceRestoreInCtxtDescForVCS is required, it is only
safe to send new contexts if the last reported event is "active to
idle". Otherwise the same context can fully preempt itself because
lite-restore is disabled.
drm/i915: Make updating pipe without modeset atomic.
Instead of doing a hack during primary plane commit the state
is updated during atomic evasion. It handles differences in
pipe size and the panel fitter.
This is continuing on top of Daniel's work to make faster
modesets atomic, and not yet enabled by default.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org>
[danvet:
- simplify/future-proof if ladder that Jesse spotted
- resolve conflict in pipe_config_check and don't spuriously move the
code.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
drm/i915: Use the atomic state in intel_update_primary_planes.
This function was still using the legacy state, convert it to atomic.
While we're at it, fix the FIXME too and disable the primary plane.
v2 (Daniel):
- Add FIXME explaining that update_primary_planes should soon get
removed anyway.
- Don't call ->disable_plane since we can't disable the primary plane
with a CS flip (noticed by Ville).
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> (v1) Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
drm/i915: add attached connector to hdmi container
This patch adds the intel_connector initialized to intel_hdmi
display, during the init phase, just like the other encoders do.
This attachment is very useful when we need to extract the connector
pointer during the hotplug handler function
Jani Nikula [Fri, 4 Sep 2015 13:55:15 +0000 (16:55 +0300)]
drm/i915: don't hard code vlv backlight frequency if unset
Fall back to VBT based backlight modulation frequency if it's not
set. Do not hard code.
This could be a problem if there is no VBT.
Cc: Clint Taylor <clinton.a.taylor@intel.com> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Tested-by: Clint Taylor <Clinton.A.Taylor@intel.com> Reviewed-by: Clint Taylor <Clinton.A.Taylor@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Jani Nikula [Fri, 4 Sep 2015 13:55:14 +0000 (16:55 +0300)]
drm/i915: initialize backlight max from VBT
Normally we determine the backlight PWM modulation frequency (which we
also use as backlight max value) from the backlight registers at module
load time, expecting the registers have been initialized by the BIOS. If
this is not the case, we fail.
The VBT contains the backlight modulation frequency in Hz. Add platform
specific functions to convert the frequency in Hz to backlight PWM
modulation frequency, and use them to initialize the backlight when the
registers are not initialized by the BIOS.
v2: Fix SPT and VLV. Thanks to Clint for the VLV code.
Cc: Clint Taylor <clinton.a.taylor@intel.com> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Reviewed-by: Clint Taylor <Clinton.A.Taylor@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Jani Nikula [Fri, 4 Sep 2015 13:55:13 +0000 (16:55 +0300)]
drm/i915: use pch backlight override on hsw too
Currently the difference between backlight control on HSW vs. BDW/SKL is
that on HSW we modify the duty cycle on the CPU register, and have the
hardware pass the changes on to the PCH registers. We still drive the
PCH PWM on both. While HSW and BDW use the same LPT PCH, BDW does not
pass these messages on to the PCH. Therefore on BDW we need to enable
the PCH override bit, and program the PCH directly. (On SPT PCH, this
mode is the default.) We could as well do this on HSW too, and in fact
I've been told this is what a certain other operating system does. So
use PCH backlight override on HSW too.
This simplifies some follow-up code, but it does have the danger of
breaking backlight on HSW machines. It should work, but mysterious are
the ways of backlight.
While at it, name the related backlight hooks according to the PCH
rather than the CPU for clarity.
Cc: Clint Taylor <clinton.a.taylor@intel.com> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Reviewed-by: Clint Taylor <Clinton.A.Taylor@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Nick Hoath [Mon, 29 Jun 2015 13:07:32 +0000 (14:07 +0100)]
drm/i915/bxt: Clean up bxt_init_clock_gating
Add stepping check for A0 workarounds, and remove the associated
FIXME tags.
Split out unrelated WAs for later condition checking.
v2: Fixed format (PeterL)
v3: Corrected stepping check for WaDisableSDEUnitClockGating
- Ignoring comment, following hardware spec instead. (ChrisH)
Added description for TILECTL setting (JonB)
Cc: Peter Lawthers <peter.lawthers@intel.com> Cc: Chris Harris <chris.harris@intel.com> Cc: Jon Bloomfield <jon.bloomfield@intel.com> Signed-off-by: Nick Hoath <nicholas.hoath@intel.com> Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
drm/i915: Change SRM, LRM instructions to use correct length
which forgot to account for the length bias when declaring the fixed
length.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=91844 Reported-by: Andreas Reis <andreas.reis@gmail.com> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Dave Gordon <david.s.gordon@intel.com> Cc: Arun Siluvery <arun.siluvery@linux.intel.com> Cc: Mika Kuoppala <mika.kuoppala@intel.com> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Reviewed-by: Arun Siluvery <arun.siluvery@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Imre Deak [Thu, 3 Sep 2015 13:24:36 +0000 (16:24 +0300)]
drm/i915: access the PP_ON_DELAYS/PP_OFF_DELAYS regs only pre GEN5
These registers exist only before GEN5, so currently we may access
undefined registers on VLV/CHV and BXT. Apply the workaround only pre
GEN5.
Since the workaround is relevant only when LVDS is present, for clarity
apply it only if this is the case.
This triggered an unclaimed register access warning on BXT.
v2: (Ville)
- move the workaround to the LVDS init code
- print a debug note about the workaround
Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Imre Deak [Thu, 3 Sep 2015 13:24:35 +0000 (16:24 +0300)]
drm/i915: access the PP_CONTROL reg only pre GEN5
This register exists only pre GEN5, but atm we also access it on
VLV/BXT/CHV. Prevent accessing it on these latter platforms.
Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Chris Wilson [Thu, 3 Sep 2015 12:01:39 +0000 (13:01 +0100)]
drm/i915: Refactor common ringbuffer allocation code
A small, very small, step to sharing the duplicate code between
execlists and legacy submission engines, starting with the ringbuffer
allocation code.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Arun Siluvery <arun.siluvery@linux.intel.com> Cc: Mika Kuoppala <mika.kuoppala@intel.com> Cc: Dave Gordon <david.s.gordon@intel.com> Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Jani Nikula [Thu, 3 Sep 2015 08:16:09 +0000 (11:16 +0300)]
drm/i915: use the yesno helper for logging
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Jani Nikula [Thu, 3 Sep 2015 08:16:08 +0000 (11:16 +0300)]
drm/i915: ignore link rate in TPS3 selection
TPS3 is mandatory for downstream devices that support HBR2, and Intel
platforms that support HBR2 also support TPS3. Whenever TPS3 is
supported by both the source and sink, it should be used. In other
words, whenever the source and sink are capable of 5.4 Gbps link, we
should anyway go for TPS3, regardless of the link rate being selected.
Log an error if the sink has advertized HBR2 capability without TPS3
capability.
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Jani Nikula [Thu, 3 Sep 2015 08:16:07 +0000 (11:16 +0300)]
drm/i915/dp: move TPS3 logic to where it's used
There is no need to have a separate flag for tps3 as the information is
only used at one location. Move the logic there to make it easier to
follow.
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Unless future specs tells otherwise we can assume future gens
inherit some stuff from the previous so let's handle
missed cases when we know tehy should't be there and assume
default equals newest one.
No functional changes.
v2: Remove useless case as pointed out by Ville.
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
drm/i915/gtt: Avoid calling kcalloc in a loop when allocating temp bitmaps
On each call to gen8_alloc_va_range_3lvl we're allocating temporary
bitmaps needed for error handling. Unfortunately, when we increase
address space size (48b ppgtt) we do additional (512 - 4) calls to
kcalloc, increasing latency between exec and actual start of execution
on the GPU. Let's just do a single kcalloc, we can also drop the size
from free_gen8_temp_bitmaps since it's no longer used.
v2: Use GFP_TEMPORARY to make the allocations reclaimable.
v3: Drop the 2D array, just allocate a single block.
v4: Rebase to handle gen8_preallocate_top_level_pdps.
v5: Align misaligned bracket.
Cc: Chris Wilson <chris@chris-wilson.co.uk> Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Cc: Michel Thierry <michel.thierry@intel.com> Signed-off-by: Michał Winiarski <michal.winiarski@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
[danvet: Correct kcalloc arguments as suggested by Chris.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
drm/i915/skl: Avoid using un-initialized bits_per_pixel
Don't rely on fb->bits_per_pixel as intel_framebuffer_init is not
filling bits_per_pixel field of fb-struct for YUV pixel format.
This leads to divide by zero error during watermark calculation.
Ville Syrjälä [Fri, 28 Aug 2015 19:59:08 +0000 (22:59 +0300)]
drm/i915: Don't call intel_get_hpd_pins() when there's no hotplug interrupt
On GMCH plaforms we are now getting the following spew on aux
interrupts:
[drm:intel_get_hpd_pins] hotplug event received, stat 0x00000000, dig 0x00000000, pins 0x00000000
[drm:intel_get_hpd_pins] hotplug event received, stat 0x00000000, dig 0x00000000, pins 0x00000000
[drm:intel_get_hpd_pins] hotplug event received, stat 0x00000000, dig 0x00000000, pins 0x00000000
[drm:intel_get_hpd_pins] hotplug event received, stat 0x00000000, dig 0x00000000, pins 0x00000000
[drm:intel_get_hpd_pins] hotplug event received, stat 0x00000000, dig 0x00000000, pins 0x00000000
[drm:intel_dp_aux_ch] dp_aux_ch timeout status 0x71450064
Prevent it by not calling intel_get_hpd_pins() unless one of the HPD
interrupt bits are actually set.
due to pch_get_hpd_pins() being chosen over i9xx_get_hpd_pins() to
serve as the new unified piece of code. pch_get_hpd_pins() had the debug
print, and i9xx_get_hpd_pins() didn't.
Cc: Imre Deak <imre.deak@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Thu, 27 Aug 2015 20:56:11 +0000 (23:56 +0300)]
drm/i915: Rewrite BXT HPD code to conform to pre-existing style
Rewrite the BXT hpd setup to match the way we do it on other platforms:
- Throw out BXT_HOTPLUG_CTL since it's the same as PCH_PORT_HOTPLUG
- Enable the HPD bits in the DE port IER in gen8_de_irq_postinstall()
- Update DE port IMR using bdw_update_port_irq()
Also throw out port D from bxt_port_hotplug_long_detect() since BXT only
goes up to C.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Thu, 27 Aug 2015 20:56:10 +0000 (23:56 +0300)]
drm/i915: Refactor the hpd irq handling functions
A lot of the hpd irq handling is duplicated code, so refactor it a bit
by observing that in several places the only difference is the hpd[]
array. So pull the code to a few functions and pass in the hpd[] array
from the caller. Another option would be to determine the correct array
to use within the functions themselves, but somehow passing it in felt
nicer.
Further code reduction could be achieved by passing in the hotplug
register offset, and the long pulse detection function pointer. But that
didn't feel as good for some reason, so I left it at the middle ground.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Thu, 27 Aug 2015 20:56:09 +0000 (23:56 +0300)]
drm/i915: Rewrite bxt_hpd_handler() to look like everyone else
bxt_hpd_handler() looks different to everyone else for no good reason.
Rewrite it to use the standard variable namees etc.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Thu, 27 Aug 2015 20:56:08 +0000 (23:56 +0300)]
drm/i915: Reinitialize HPD after runtime D3
Runtime suspends disabled all interrupts, so in order to get them back
fully we need to also do the HPD irq setup on runtime resume. Except
on VLV/CHV where the display interrupt initialization is part of the
display power well powerup.
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Thu, 27 Aug 2015 20:56:07 +0000 (23:56 +0300)]
drm/i915: Add port A HPD support for SPT
On SKL the port A HPD has moved to the PCH. Hook it up.
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Thu, 27 Aug 2015 20:56:06 +0000 (23:56 +0300)]
drm/i915: Add port A HPD support for BDW
Wire up the port A HPD for BDW. Compared to earlier platforms the
interrupt setup is a bit different, but basically everything else
looks the same.
v2: 0 initialize pin_mask/long_mask due to intel_get_hpd_pins() changes
Check for BDW before processing the HPD to not break BXT
Set found=true when processing port A HPD
Sort out the mess I made of the irq setup in v1
Warn about bad irq mask vs. enable bits in bdw_update_port_irq() (Paulo)
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Thu, 27 Aug 2015 20:56:05 +0000 (23:56 +0300)]
drm/i915: LPT:LP needs port A HPD enabled in both north and south
If the CPU and PCH are on the same package we must enabled the port A
HPD also in the south hotplug register. To identify the package type
we simply look at the PCH type: LPT-H means separate package, and
LPT-LP means multi chip package (MCP).
v2: Add comment and pimp commit message
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Thu, 27 Aug 2015 20:56:04 +0000 (23:56 +0300)]
drm/i915: Add port A HPD support for IVB/HSW
As with ILK/SNB wire up the port A HPD on IVB/HSW.
This might be more important on HSW with PSR. BSpec tells us that if the
automagic link training performed by the hardware fails for some reason,
we're going to get a short HPD and are supposed to re-train the link
manyally.
v2: 0 initialize pin_mask/long_mask due to intel_get_hpd_pins() changes
Add a comment about the pulse duration bits being reserved on HSW+
like we have for LPT+ in ibx_hpd_irq_setup()
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Thu, 27 Aug 2015 20:56:03 +0000 (23:56 +0300)]
drm/i915: Add port A HPD support for ILK/SNB
ILK/SNB support port A HPD. While HPD is optional on eDP let's at least
try to wite it up so that we might notice if the link has issues.
The eDP spec suggests that if HPD is not wired up, one should poll the
link status instead. We don't even do that currently.
v2: 0 initialize pin_mask/long_mask due to intel_get_hpd_pins() changes
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Thu, 27 Aug 2015 20:56:02 +0000 (23:56 +0300)]
drm/i915: Introduce spt_irq_handler()
Starting from SPT the only interrupts living in the south are GMBUS and
HPD. What's worse some of the SPT specific new bits conflict with some
other bits on earlier PCH generations. So better not use the
cpt_irq_handler() for SPT+ anymore.
Also kill the hand rolled port E handling with something more
standardish. This also avoids accidentally confusing port B and port E
long pulses since the bits occupy the same positions, just in different
registers.
Also add a comment noting that the short pulse duration bits are
reserved on LPT+. The 2ms value we program is 0, so no issue wrt. the
MBZ in the spec.
v2: Call intel_hpd_irq_handler() only once (Paulo)
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Fri, 28 Aug 2015 18:26:27 +0000 (21:26 +0300)]
drm/i915: Move {pin, long}_mask initialization to caller from intel_get_hpd_pins()
Move the 0 initialization of pin_mask and long_mask from
intel_get_hpd_pins() into each caller. This we we can call
intel_get_hpd_pins() multiple times to accumulate more pins from several
sources.
v2: Add a comment explaining the dangers of intel_get_hpd_pins() (Paulo)
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Thu, 27 Aug 2015 20:56:00 +0000 (23:56 +0300)]
drm/i915: Rename BXT PORTA HPD defines
The PORTA HPD defines are not BXT specific. They also exist on SPT,
and partially already on LPT:LP.
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Thu, 27 Aug 2015 20:55:59 +0000 (23:55 +0300)]
drm/i915: Add HAS_PCH_LPT_LP() macro
Make LPT:LP checks look neater by wrapping the details in a
new HAS_PCH_LPT_LP() macro.
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Thu, 27 Aug 2015 20:55:58 +0000 (23:55 +0300)]
drm/i915: Factor out ilk_update_display_irq()
Extract the core of ironlake_{enable,disable}_display_irq() into a new
function. We'll have further use for it later.
v2: Warn about invalid mask vs. enable bits (Paulo)
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Thu, 27 Aug 2015 20:55:57 +0000 (23:55 +0300)]
drm/i915: Extract intel_hpd_enabled_irqs()
Eliminate a bunch of duplicated code that calculates the currently
enabled HPD interrupt bits.
v2: s/;/:/ in patch subject (Paulo)
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Thu, 27 Aug 2015 20:55:56 +0000 (23:55 +0300)]
drm/i915: Clean up various HPD defines
Indent the PORTx_HOTPLUG_... defines appropriately, and fix some space
vs. tab issues.
v2: Document pre-HSW/LPT bits, and order another tab (Paulo)
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Zhiyuan Lv [Fri, 28 Aug 2015 07:41:18 +0000 (15:41 +0800)]
drm/i915: guest i915 notification for Intel GVT-g
When i915 drivers run inside a VM with Intel GVT-g, some explicit
notifications are needed from guest to host device model through PV
INFO page write. The notifications include:
PPGTT create
PPGTT destroy
They are used for the shadow implementation of PPGTT. Intel GVT-g
needs to write-protect the guest pages of PPGTT, and clear the write
protection when they end their life cycle.
v2:
- Use lower_32_bits()/upper_32_bits() for qword operations;
- Remove the notification of guest context creation/destroy;
Signed-off-by: Zhiyuan Lv <zhiyuan.lv@intel.com> Signed-off-by: Zhi Wang <zhi.a.wang@intel.com> Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Zhiyuan Lv [Fri, 28 Aug 2015 07:41:17 +0000 (15:41 +0800)]
drm/i915: Update PV INFO page definition for Intel GVT-g
Some more definitions in the PV info page are added. They are mainly
for the guest notification to Intel GVT-g device model. They are used
for Broadwell enabling.
The notification of PPGTT page table creation/destroy is to notify
GVT-g device model the life cycle of guest page tables. Then device
model will implement shadow page table for guests.
The notification of context create/destroy is optional. If it is used,
the device model will create/destroy shadow context corresponding to
the context's life cycle. Guest driver needs to make sure that the
context's LRCA and backing storage address unchanged. If it is not
used, the device model will perform the context shadow work in the
context scheduling time.
Signed-off-by: Zhiyuan Lv <zhiyuan.lv@intel.com> Signed-off-by: Zhi Wang <zhi.a.wang@intel.com> Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Zhiyuan Lv [Fri, 28 Aug 2015 07:41:16 +0000 (15:41 +0800)]
drm/i915: Always enable execlists on BDW for vgpu
Broadwell hardware supports both ring buffer mode and execlist mode.
When i915 runs inside a VM with Intel GVT-g, we allow execlist mode
only.
The main reason of EXECLIST only is that GVT-g does not support the
dynamic mode switch between ring buffer mode and execlist mode when
running multiple virtual machines.
v2:
- Adjust the position of vgpu check in sanitize function (Joonas)
- Add vgpu error check in context initialization. (Joonas, Daniel)
Signed-off-by: Zhiyuan Lv <zhiyuan.lv@intel.com> Signed-off-by: Zhi Wang <zhi.a.wang@intel.com> Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Zhiyuan Lv [Fri, 28 Aug 2015 07:41:14 +0000 (15:41 +0800)]
drm/i915: preallocate pdps for 32 bit vgpu
This is based on Mika Kuoppala's patch below:
http://article.gmane.org/gmane.comp.freedesktop.xorg.drivers.intel/61104/match=workaround+hw+preload
The patch will preallocate the page directories for 32-bit PPGTT when
i915 runs inside a virtual machine with Intel GVT-g. With this change,
the root pointers in EXECLIST context will always keep the same.
The change is needed for vGPU because Intel GVT-g will do page table
shadowing, and needs to track all the page table changes from guest
i915 driver. However, if guest PPGTT is modified through GPU commands
like LRI, it is not possible to trap the operations in the right time,
so it will be hard to make shadow PPGTT to work correctly.
Shadow PPGTT could be much simpler with this change. Meanwhile
hypervisor could simply prohibit any attempt of PPGTT modification
through GPU command for security.
The function gen8_preallocate_top_level_pdps() in the patch is from
Mika, with only one change to set "used_pdpes" to avoid duplicated
allocation later.
Cc: Mika Kuoppala <mika.kuoppala@intel.com> Cc: Dave Gordon <david.s.gordon@intel.com> Cc: Michel Thierry <michel.thierry@intel.com> Signed-off-by: Zhiyuan Lv <zhiyuan.lv@intel.com> Signed-off-by: Zhi Wang <zhi.a.wang@intel.com> Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Jani Nikula [Thu, 27 Aug 2015 13:23:30 +0000 (16:23 +0300)]
drm/i915: add yesno utility function
Add a common function to return "yes" or "no" string based on the
argument, and drop the local versions of it.
Signed-off-by: Jani Nikula <jani.nikula@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Jani Nikula [Wed, 26 Aug 2015 07:58:20 +0000 (10:58 +0300)]
drm/i915: move intel_hrawclk() to intel_display.c
Make it available outside of intel_dp.c.
Signed-off-by: Jani Nikula <jani.nikula@intel.com> Reviewed-by: Clint Taylor <Clinton.A.Taylor@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Alex Dai [Tue, 18 Aug 2015 21:32:35 +0000 (14:32 -0700)]
drm/i915/guc: Support GuC version 4.3
The firmware layout changes that now it only has css header +
uCode + RSA signature. Plus, other trivial changes to support
GuC V4.3.
Signed-off-by: Alex Dai <yu.dai@intel.com> Reviewed-by: Dave Gordon <david.s.gordon@intel.com> Reviewed-by: Dave Gordon <david.s.gordon@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The driver doesn't support UMS any more, so set DRIVER_MODESET by default,
remove the legacy s/r callbacks, and rename the s/r functions to make it more clear
they're only in use by switcheroo now.
Also remove an obsolete comment about atomic. Normal updates are supported only
async updates aren't yet.
v2: Don't unconditionally set DRIVER_ATOMIC, we're not yet there.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Wed, 26 Aug 2015 16:39:18 +0000 (19:39 +0300)]
drm/i915: Fix clock readout when pipes are enabled w/o ports
The BIOS sometimes likes to enable pipes w/o any ports, at least on
older machines. Currently we fail to assign anything sensible to
crtc->hwmode.crtc_clock which leads to complaints from the vblank code.
Deal with active pipes w/o ports and assign something sensible to
crtc_clock in i9xx_get_pipe_config(). The encoder .get_config() will
override this if the port is enabled.
Gets rid of rest of these on my gen4:
[drm:drm_calc_timestamping_constants [drm]] *ERROR* crtc 24: Can't calculate constants, dotclock = 0!
[drm:i915_get_vblank_timestamp] crtc 1 is disabled
v2: Fill out crtc_clock already in i9xx_get_pipe_config() (Maarten)
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Wed, 8 Jul 2015 20:46:01 +0000 (23:46 +0300)]
drm/i915: Add CHV PHY LDO power sanity checks
At various points when changing the DPIO lane/phy power states,
construct an expected value of the DISPLAY_PHY_STATUS register
and compare it with the real thing.
To construct the expected value we look at our shadow PHY_CONTROL
register value (which should match what we've just written to the
hardware), and we also need to look at the actual state of the cmn
power wells as a disabled power well causes the relevant LDO status
to be reported as 'on' in DISPLAY_PHY_STATUS.
When initially powering up the PHY it performs various internal
calibrations for which it fully powers up. That means that if we check
for the expetected power state immediately upon releasing cmnreset we
would get the occasional false positive. But we can of course
poll until the expected value appears. It shouldn't be too long so
this shouldn't make modesets substantially longer.
One extra complication is introduced when we cross the streams, ie.
drive port B with pipe B. In this case we trick CL2 (where the DPLL lives)
into life by temporaily powering up the lanes in the second channel,
and once the pipe is up and runnign we release the lane power override.
At that point the power state of CL2 has somehow gotten entangled with
the power state of the first channel. That means that constructing the
expected DISPLAY_PHY_STATUS value is a bit tricky since based on the
lane power states in the second channel, CL2 should also be powered
down. But we can use the DPLL enable bit to determine when CL2 should
be alive even if the lanes are powered down. However the power state
of CL2 isn't actually tied in with the DPLL state, but to the state
of the lanes in first channel, so we have to avoid checking the
expected state between shutting down the DPLL and powering down
the lanes in the first channel. So no calling assert_chv_phy_status()
before the DISPLAY_PHY_CONTROL write in chv_phy_powergate_lanes(),
but after the write is a safe time to check.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Deepak S <deepak.s@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Wed, 8 Jul 2015 20:46:00 +0000 (23:46 +0300)]
drm/i915: Add some CHV DPIO lane power state asserts
Add some checks that the state of the DPIO lanes is more or less what we
expect based on the overrides.
The hardware only provides two bits per channel indicating whether all
or some of the lanes are powered down, so we can't do an exact check.
Additionally, CL2 powering down before we can check it adds another
twist. To work around this we simply check for the 0 value of the
CL2 register (which is what we get when it's powered down) and
adjust our expectations.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Deepak S <deepak.s@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Thu, 9 Jul 2015 17:14:11 +0000 (20:14 +0300)]
drm/i915: Clean up CHV lane soft reset programming
Currently we release the lane soft reset before lane stagger settings
have been programmed. I believe that means we don't actually do lane
staggering. So move the soft reset deassert to happen after lane
staggering has been programmed.
The one confusing thing in this is that when we remove the power down
override from the lanes, they power up with defaul register values,
which do not have the soft reset overrides enabled. And according to
some docs by default the data lane resets are tied to cmnreset. So that
would mean that lanes would come out of reset without staggering as
soon as the power down overrides are removed. But since we can't access
either the lane stagger register nor the soft reset override registers
until the lanes are powered on, we can't really do anything about it.
So let's just set the soft reset overrides as soon as the lane is
powered on and hope for the best.
v2: Fix typos in commit message (Daniel)
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Deepak S <deepak.s@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The DP MST encoder config function never sets ddi_pll_sel, even though
its value is programmed in its ->pre_enable() hook. That used to work
because a new pipe_config was kzalloc'ed at every modeset, and the value
of zero selects the highest clock for the PLL. Starting with the commit
below, the value of ddi_pll_sel is preserved through modesets, and since
the correct value wasn't properly setup by the MST code, it could lead
to warnings and blank screens.
commit 8504c74c7ae48b4b8ed1f1c0acf67482a7f45c93
Author: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Date: Fri May 15 11:51:50 2015 +0300
drm/i915: Preserve ddi_pll_sel when allocating new pipe_config
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=91628 Cc: stable@vger.kernel.org # 7e6313a2516d drm/i915: Don't use link_bw for PLL setup Cc: stable@vger.kernel.org Cc: Timo Aaltonen <tjaalton@ubuntu.com> Cc: Luciano Coelho <luciano.coelho@intel.com> Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Jani Nikula [Thu, 27 Aug 2015 10:25:37 +0000 (13:25 +0300)]
drm/i915/dp: use the drm dp helper for determining sink tps3 support
No functional changes.
Signed-off-by: Jani Nikula <jani.nikula@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
[danvet: s/intel_dp_tps/drm_dp_tps/.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Tue, 11 Aug 2015 17:21:46 +0000 (20:21 +0300)]
drm/i915: Don't use link_bw for PLL setup
Use port_clock instead of link_bw when picking the PLL parameters for
DP. link_bw may be zero with an eDP 1.4 sink that supports
DP_LINK_RATE_SET so we shouldn't use it for anything other than feed it
to the sink appropriately.
v2: Fix typo in commit message (Sivakumar)
Reviewed-by: Sivakumar Thulasimani <sivakumar.thulasimani@intel.com> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
[Jani: cherry-picked from future.] Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Commit 92122789b2d6 ("drm/i915: preserve SSC if previously set v3")
added code to intel_modeset_gem_init to override the SSC status read
from VBT with the SSC status set by BIOS.
However, intel_modeset_gem_init is invoked *after* intel_modeset_init,
which calls intel_setup_outputs, which *modifies* SSC status by way of
intel_init_pch_refclk. So unlike advertised, intel_modeset_gem_init
doesn't preserve the SSC status set by BIOS but whatever
intel_init_pch_refclk decided on.
This is a problem on dual gpu laptops such as the MacBook Pro which
require either a handler to switch DDC lines, or the discrete gpu
to proxy DDC/AUX communication: Both the handler and the discrete
gpu may initialize after the i915 driver, and consequently, an LVDS
connector may initially seem disconnected and the SSC therefore
is disabled by intel_init_pch_refclk, but on reprobe the connector
may turn out to be connected and the SSC must then be enabled.
Due to 92122789b2d6 however, the SSC is not enabled on reprobe since
it is assumed BIOS disabled it while in fact it was disabled by
intel_init_pch_refclk.
Also, because the SSC status is preserved so late, the preserved value
only ever gets used on resume but not on panel initialization:
intel_modeset_init calls intel_init_display which indirectly calls
intel_panel_use_ssc via multiple subroutines, *before* the BIOS value
overrides the VBT value in intel_modeset_gem_init (intel_panel_use_ssc
is the sole user of dev_priv->vbt.lvds_use_ssc).
Fix this by moving the code introduced by 92122789b2d6 from
intel_modeset_gem_init to intel_modeset_init before the invocation
of intel_setup_outputs and intel_init_display.
Add a DRM_DEBUG_KMS as suggested way back by Jani:
http://lists.freedesktop.org/archives/intel-gfx/2014-June/046666.html
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=88861
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=61115 Tested-by: Paul Hordiienko <pvt.gord@gmail.com>
[MBP 6,2 2010 intel ILK + nvidia GT216 pre-retina] Tested-by: William Brown <william@blackhats.net.au>
[MBP 8,2 2011 intel SNB + amd turks pre-retina] Tested-by: Lukas Wunner <lukas@wunner.de>
[MBP 9,1 2012 intel IVB + nvidia GK107 pre-retina] Tested-by: Bruno Bierbaumer <bruno@bierbaumer.net>
[MBP 11,3 2013 intel HSW + nvidia GK107 retina -- work in progress] Fixes: 92122789b2d6 ("drm/i915: preserve SSC if previously set v3") Signed-off-by: Lukas Wunner <lukas@wunner.de> Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org> Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Rodrigo Vivi [Sat, 8 Aug 2015 00:35:21 +0000 (17:35 -0700)]
drm/i915/skl: Enable DDI-E
There are OEMs using DDI-E out there,
so let's enable it.
Unfortunately there is no detection bit for DDI-E
So we need to rely on VBT for that.
I also need to give credits to Xiong since before seing
his approach to check info->support_* I was creating an ugly
vbt->ddie_sfuse_strap in order to propagate the ddi presence info
v2: Rebased as last patch in the series. since all other patches
in this series are needed for anything working propperly on DDI-E.
drm/i915: gen4: work around hang during hibernation
using an explicit blacklist for the GENs/BIOS vendors where the issue was
reported. Later there we had reports of the same failure on platforms not on
this list.
To my best knowledge the correct thing to do is still to put the device to PCI
D3 state during hibernation, see [1] and [2] for the reasons. This also aligns
with our future plans to unify more the runtime and system suspend/resume
paths. Since an exact blacklist seems to be impractical (multiple GENs and
BIOS vendors are affected) apply the workaround on everything pre GEN6.
Ville Syrjälä [Thu, 20 Aug 2015 16:37:29 +0000 (19:37 +0300)]
drm/i915: Check DP link status on long hpd too
We are no longer checkling the DP link status on long hpd. We used to do
that from the .hot_plug() handler, but it was removed when MST got
introduced.
If there's no userspace we now fail to retrain the link if the sink
power is toggled (or cable yanked and replugged), meaning the user is
left staring at a blank screen. With the retraining put back that should
be fixed.
Also remove the leftover comment that referred to the old retraining
from .hot_plug().
Gary Wang [Fri, 28 Aug 2015 08:40:34 +0000 (16:40 +0800)]
drm/i915: set CDCLK if DPLL0 enabled during resuming from S3
Since BIOS RC 1.4 it would enable CDCLK PLL during BIOS S3 resume, then
driver needs to set CDCLK to avoid display corruption if DPLL0 enabled.
References: https://bugs.freedesktop.org/show_bug.cgi?id=91697 Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Reviewed-by: Damien Lespiau <damien.lespiau@intel.com> Reviewed-by: Cooper Chiou <cooper.chiou@intel.com> Reviewed-by: Wei Shun Chang <wei.shun.chang@intel.com> Tested-by: Gary Wang <gary.c.wang@intel.com> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: Gavin Hindman <gavin.hindman@intel.com> Cc: Chris Wilson <chris@chris-wilson.co.uk> Cc: Xiong Y Zhang <xiong.y.zhang@intel.com> Signed-off-by: Gary Wang <gary.c.wang@intel.com> Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Paulo Zanoni [Tue, 25 Aug 2015 22:03:41 +0000 (19:03 -0300)]
drm/i915: gen 9 can check for unclaimed registers too
Dear git bisect user,
Even though this is the patch that introduced the WARN() you're
bisecting, please notice that it's very likely that the problem you're
facing was already present before this commit. In other words: this
commit adds code to detect errors and give WARN()s about them, but the
errors were already there.
In order to continue your debug, please use the i915.mmio_debug
option, check the backtraces and try to discover which read or write
operation is causing the error message. Then check if this is
happening because the register does not exist or because its power
well is down when the operation is being done.
On my SKL machine, if I use i915.mmio_debug=999, this patch triggers
42 WARNs just by booting. I didn't investigate them yet. Normal users
are only going to get a single WARN due to the default i915.mmio_debug
setting.
Thank you for your comprehension,
Paulo
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Wed, 8 Jul 2015 20:45:58 +0000 (23:45 +0300)]
drm/i915: Force CL2 off in CHV x1 PHY
We can choose to leave the display PHY CL2 powerdown up to some hardware
signals, or we can force it. The BXT code forces the nonexistent CL2 in
the x1 PHY to power down. Follow suit on CHV. Maybe it can still save
some extra power by disabling some extra logic in CL1, or something.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Deepak S <deepak.s@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Wed, 8 Jul 2015 20:45:57 +0000 (23:45 +0300)]
drm/i915: Enable DPIO SUS clock gating on CHV
CHV has supports some form of automagic clock gating for the
DPIO SUS clock. We can simply enable the magic bits and the
hardware should take care of the rest.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Deepak S <deepak.s@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Fri, 10 Jul 2015 07:56:24 +0000 (10:56 +0300)]
drm/i915: Force common lane on for the PPS kick on CHV
With DPIO powergating active the DPLL can't be accessed unless
something else is keeping the common lane in the channel on.
That means the PPS kick procedure could fail to enable the PLL.
Power up some data lanes to force the common lane to power up
so that the PLL can be enabled temporarily.
v2: Avoid gcc uninitilized variable warning
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Deepak S <deepak.s@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Wed, 8 Jul 2015 20:45:55 +0000 (23:45 +0300)]
drm/i915: Trick CL2 into life on CHV when using pipe B with port B
Normmally the common lane in a PHY channel gets powered up when some
of the data lanes get powered up. But when we're driving port B with
pipe B we don't want to enabled any of the data lanes, and just want
the DPLL in the common lane to be active.
To make that happens we have to temporarily enable some data lanes
after which we can access the DPLL registers in the common lane. Once
the pipe is up and running we can drop the power override on the data
lanes allowing them to shut down. From this point forward the common
lane will in fact stay powered on until the data lanes in the other
channel get powered down.
Ville's extended explanation from the review thread:
On Wed, Aug 19, 2015 at 07:47:41AM +0530, Deepak wrote:
> One Q, why only for port B? Port C is also in same common lane right?
Port B is in the first PHY channel which also houses CL1. CL1 always
powers up whenever any lanes in either PHY channel are powered up.
CL2 only powers up if lanes in the second channel (ie. the one with
port C) powers up.
So in this scenario (pipe B->port B) we want the DPLL from CL2, but
ideally we only want to power up the lanes for port B. Powering up
port B lanes will only power up CL1, but as we need CL2 instead we
need to, temporarily, power up some lanes in port C as well.
Crossing the streams the other way (pipe A->port C) is not a problem
since CL1 powers up whenever anything else powers up. So powering up
some port C lanes is enough on its own to make the CL1 DPLL
operational, even though CL1 and the lanes live in separate channels.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Deepak S <deepak.s@linux.intel.com>
[danvet: Amend commit message with extended explanation.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Wed, 8 Jul 2015 20:45:54 +0000 (23:45 +0300)]
drm/i915: Implement PHY lane power gating for CHV
Powergate the PHY lanes when they're not needed. For HDMI all four lanes
are needed always, but for DP we can enable only the needed lanes. To
power down the unused lanes we use some power down override bits in the
DISPLAY_PHY_CONTROL register. Without the overrides it appears that the
hardware always powers on all the lanes. When the port is disabled the
power down override is not needed and the lanes will shut off on their
own. That also means the override is critical to actually be able to
access the DPIO registers before the port is actually enabled.
Additionally the common lanes will power down when not needed. CL1
remains on as long as anything else is on, CL2 will shut down when
all the lanes in the same channel will shut down. There is one exception
for CL2 that will be dealt in a separate patch for clarity.
With potentially some lanes powered down, the DP code now has to check
the number of active lanes before accessing PCS/TX registers. All
registers in powered down blocks will reads as 0xffffffff, and soe we
would drown in warnings from vlv_dpio_read() if we allowed the code
to access all those registers.
Another important detail in the DP code is the "TX latency optimal"
setting. Normally the second TX lane acts as some kind of reset master,
with the other lanes as slaves. But when only a single lane is enabled,
that single lane obviously has to be the master.
A bit of extra care is needed to reconstruct the initial state of the
DISPLAY_PHY_CONTROL register since it can't be read safely. So instead
read the actual lane status from the DPLL/PHY_STATUS registers and
use that to determine which lanes ought to be powergated initially.
We also need to switch the PHY power modes to "deep PSR" to avoid
a hard system hang when powering down the single channel PHY.
Also sprinkle a few debug prints around so that we can monitor the
DISPLAY_PHY_STATUS changes without having to read it and risk
corrupting it.
v2: Add locking to chv_powergate_phy_lanes()
v3: Actually enable dynamic powerdown in the PHY and deal with the
fallout
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Deepak S <deepak.s@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ville Syrjälä [Mon, 29 Jun 2015 12:25:53 +0000 (15:25 +0300)]
drm/i915: Move DPLL ref/cri/VGA mode frobbing to the disp2d well enable
Bunch of stuff needs the DPLL ref/cri clocks on both VLV and CHV,
and having VGA mode enabled causes some problems for CHV. So let's just
pull the code to configure those bits into the disp2d well enable hook.
With the DPLL disable code also fixed to leave those bits alone we
should now have a consistent DPLL state all the time even if the DPLL
is disabled.
This also neatly removes some duplicated code between the VLV and
CHV codepaths.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Sivakumar Thulasimani <sivakumar.thulasimani@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
projects / linux-beck.git / log