This adds DP 1.2 MST support on Haswell systems.

Notes:
a) this reworks irq handling for DP MST ports, so that we can
avoid the mode config locking in the current hpd handlers, as
we need to process up/down msgs at a better time.

Changes since v0.1:
use PORT_PCH_HOTPLUG to detect short vs long pulses
add a workqueue to deal with digital events as they can get blocked on the
main workqueue beyong mode_config mutex
fix a bunch of modeset checker warnings
acks irqs in the driver
cleanup the MST encoders

Changes since v0.2:
check irq status again in work handler
move around bring up and tear down to fix DPMS on/off
use path properties.

Changes since v0.3:
updates for mst apis
more state checker fixes
irq handling improvements
fbcon handling support
improved reference counting of link - fixes redocking.

Changes since v0.4:
handle gpu reset hpd reinit without oopsing
check link status on HPD irqs
fix suspend/resume

Changes since v0.5:
use proper functions to get max link/lane counts
fix another checker backtrace - due to connectors disappearing.
set output type in more places fro, unknown->displayport
don't talk to devices if no HPD asserted
check mst on short irqs only
check link status properly
rebase onto prepping irq changes.
drop unsued force_act

Changes since v0.6:
cleanup unused struct entry.

[airlied: fix some sparse warnings].
Reviewed-by: NTodd Previte <tprevite@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

for MST I need to reuse this, so just move it now.
Signed-off-by: NDave Airlie <airlied@redhat.com>

DP MST will need connectors that aren't connected to specific
encoders, add some checks in advance to avoid oopses.
Reviewed-by: NTodd Previte <tprevite@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

this is just prep work for mst support.
Reviewed-by: NTodd Previte <tprevite@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

These are just from the Haswell spec.
Reviewed-by: NTodd Previte <tprevite@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Not enough brown paper bags, you'll have to share one.
(oops below).

The initial match condition code was racy (locking is coming I hear).

then along came:
cd234b0b
drm/i915: Do not dereference NULL crtc or fb until after checking

Chris made an attempt to fix it, Ben "reviewed" it.
Daniel merged it.

Then
drm/i915: Make use of intel_fb_obj() (v2)
2ff8fde1

made it worse by removing the obj check later.

All in all, my laptop can't barely turn off the display
without hitting this.

Posted to #intel-gfx out of niceness, but I've merged
this already into drm-next.

Here's an oops.
[   11.528185] BUG: unable to handle kernel NULL pointer dereference at 00000000000000d0
[   11.528233] IP: [<ffffffffa0161fde>] intel_edp_psr_match_conditions+0x1e/0x2e0 [i915]
[   11.528294] PGD 35bc0067 PUD c997c067 PMD 0
[   11.528321] Oops: 0000 [#1] SMP
[   11.528916] CPU: 3 PID: 244 Comm: kworker/3:2 Not tainted 3.16.0-rc4+ #17
[   11.528949] Hardware name: LENOVO 20ARS25701/20ARS25701, BIOS GJET72WW (2.22 ) 02/21/2014
[   11.529004] Workqueue: events intel_edp_psr_work [i915]
[   11.529031] task: ffff8803079fdaa0 ti: ffff8803079c4000 task.ti: ffff8803079c4000
[   11.529067] RIP: 0010:[<ffffffffa0161fde>]  [<ffffffffa0161fde>] intel_edp_psr_match_conditions+0x1e/0x2e0 [i915]
[   11.529129] RSP: 0018:ffff8803079c7d40  EFLAGS: 00010246
[   11.529155] RAX: 0000000000000000 RBX: ffff88030c11c000 RCX: c000000000000000
[   11.529189] RDX: 0000000000000001 RSI: 1df0000000000000 RDI: ffff88030c1190d8
[   11.529222] RBP: ffff8803079c7d60 R08: ffffffff82691140 R09: 0000000000000000
[   11.529256] R10: ffff8803079fdaa0 R11: 3e00000000000000 R12: ffff88030c11c728
[   11.529290] R13: ffff88030c1190d8 R14: ffff88031e2d8e00 R15: 00000000000000c0
[   11.529324] FS:  0000000000000000(0000) GS:ffff88031e2c0000(0000) knlGS:0000000000000000
[   11.529361] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[   11.529389] CR2: 00000000000000d0 CR3: 00000000c8d9d000 CR4: 00000000001407e0
[   11.529423] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[   11.529457] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[   11.529489] Stack:
[   11.529500]  ffff88030c119000 ffff88030c11c728 ffff88030c1190d8 ffff88031e2d8e00
[   11.529541]  ffff8803079c7d88 ffffffffa01679b2 ffff880035b29a80 ffff880307909f00
[   11.529583]  ffff88031e2d4740 ffff8803079c7df8 ffffffff810a78ab ffffffff810a7849
[   11.529624] Call Trace:
[   11.529654]  [<ffffffffa01679b2>] intel_edp_psr_work+0x52/0x90 [i915]
[   11.529689]  [<ffffffff810a78ab>] process_one_work+0x1db/0x540
[   11.529719]  [<ffffffff810a7849>] ? process_one_work+0x179/0x540
[   11.529750]  [<ffffffff810a81ed>] worker_thread+0x11d/0x520
[   11.529779]  [<ffffffff810a80d0>] ? create_and_start_worker+0x60/0x60
[   11.529810]  [<ffffffff810aeb04>] kthread+0xe4/0x100
[   11.529836]  [<ffffffff810aea20>] ? kthread_create_on_node+0x200/0x200
[   11.529870]  [<ffffffff81705ebc>] ret_from_fork+0x7c/0xb0
[   11.529896]  [<ffffffff810aea20>] ? kthread_create_on_node+0x200/0x200
[   11.529926] Code: ba 31 13 f0 c9 85 f6 75 84 eb d0 66 90 0f 1f 44 00 00 55 48 89 e5 41 56 41 55 41 54 53 48 8b 87 68 ff ff ff 48 8b 9f 28 ff ff ff <48> 8b 80 d0 00 00 00 4c 8b 63 28 48 8b 40 48 48 85 c0 0f 84 1a
[   11.530110] RIP  [<ffffffffa0161fde>] intel_edp_psr_match_conditions+0x1e/0x2e0 [i915]
[   11.530163]  RSP <ffff8803079c7d40>
[   11.530180] CR2: 00000000000000d0
Signed-off-by: NDave Airlie <airlied@redhat.com>

Signed-off-by: NRob Clark <robdclark@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

We can eliminate a lot of special case code by making the computation of
the interrupt mask be correct for all callers.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Place the RPS counters inside the RPS struct.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

C is super happy to asign anything pointer to void *. Don't pretend
otherwise.
Signed-off-by: NDamien Lespiau <damien.lespiau@intel.com>
Reviewed-by: NPaulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

We may reach this point while the machine is still runtime suspended,
so we'll hit a WARN. The other encoders also don't touch registers at
this point, so instead of waking the machine up, write some code to
keep the register always at the same state, including after we runtime
suspend/resume.

Testcase: igt/pm_rpm
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=80463Signed-off-by: NPaulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: NDamien Lespiau <damien.lespiau@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Just like we do for the other encoders. This should fix some WARNs
when running pm_rpm on SNB.

Testcase: igt/pm_rpm
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=80463Signed-off-by: NPaulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: NDamien Lespiau <damien.lespiau@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Just like we already do in haswell_get_pipe_config(). This should
prevent some WARNs when we run pm_rpm on SNB.

Testcase: igt/pm_rpm
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=80463Signed-off-by: NPaulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: NDamien Lespiau <damien.lespiau@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Since we now have support for shared DPLLS.
Signed-off-by: NPaulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: NDamien Lespiau <damien.lespiau@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

To be able to do this we need to separately keep track of how many
crtcs need a given WRPLL and how many actually actively use it. The
common shared dpll framework already has all this, including massive
state readout and cross checking. Which allows us to do this switch in
a fairly small patch.
Reviewed-by: NDamien Lespiau <damien.lespiau@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Mostly this patch is one big excersize in deleting code and asserts
which are no longer needed. Note that we still abuse the shared dpll
framework a bit since we call the enable/disable functions from the
crtc mode_set and off hooks. But changing the actual hardware sequence
will be done in the next step.

Note that besides the massive amount of changes in this patch the
places and order in which the low-level WRPLL code is called is
absolutely unchanged.
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: NDamien Lespiau <damien.lespiau@intel.com>
[imre: rebased on patchset version w/o pch/crt/fdi refactoring]
Signed-off-by: NImre Deak <imre.deak@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

This time around another cute hack to pre-fill the pll->hw_state with
the right values. And also remove a bunch of checks which will be
replaced by lots more checks in the common framework.
Reviewed-by: NDamien Lespiau <damien.lespiau@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Currently still with a redudant WARN_ON in there, the common shared
dpll code will take care of this in the future.

Also we need to flip the switch for the transitional hack now to make
sure that we disable the right pll.
Reviewed-by: NDamien Lespiau <damien.lespiau@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Still tacked onto the side, but slowly getting there.

v2: Don't forget the debugfs file.

v3 (from Paulo): Don't forget to check the power domains.
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: NPaulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: NDamien Lespiau <damien.lespiau@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

And get/put it when needed. The special thing about this commit is
that it will now return false in ibx_pch_dpll_get_hw_state() in case
the power domain is not enabled. This will fix some WARNs we have when
we run pm_rpm on SNB.

Testcase: igt/pm_rpm
Bugzilla:https://bugs.freedesktop.org/show_bug.cgi?id=80463Signed-off-by: NPaulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: NDamien Lespiau <damien.lespiau@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

The WRPLLs won't use them.
Reviewed-by: NDamien Lespiau <damien.lespiau@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Just filing in names and ids, but not yet officially registering them
so that the hw state cross checker doesn't completely freak out about
them. Still since we do already read out and cross check
config->shared_dpll the basics are now there to flesh out the wrpll
shared dpll implementation.

The idea is now to roll out all the callbacks step-by-step and then at
the end switch to the shared dpll framework. This way hw and sw
changes are clearly separated.
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
[imre: added const to hsw_ddi_pll_names (Damien)]
Reviewed-by: NDamien Lespiau <damien.lespiau@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

This way only the dynamic WRPLL selection for hdmi ddi mode is
done in intel_ddi_pll_select.

v2: Don't clobber the precomputed values when selecting clocks fro
hdmi encoders.
v3 (from Paulo): Rebase on top of the s/IS_HASWELL/HAS_DDI/ patch.
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: NPaulo Zanoni <przanoni@gmail.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Don't let it fall in the HAS_PCH_SPLIT() case.
Signed-off-by: NPaulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: NDamien Lespiau <damien.lespiau@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

To make things a bit more manageable extract a new function for
reading out common ddi port state. This means a bit of duplication
between encoders and the core since both look at the same registers,
but doesn't seem worth to make a fuzz about.

We can also remove the state readout code in intel_ddi_setup_hw_pll_state.
That code is only called from the hardware take over and not the cross
check code, and only after the crtc state is reconstructed. So we can
rely on an accurate value of crtc->config.ddi_pll_sel already.

Compared to the old code also trust the hw state more and don't
special-case port A - we want to cross-check the actual-state, not
bake in our own assumptions about how this is supposed to all be
linked up.

v2: Make use of the read-out ddi_pll_sel in intel_ddi_clock_get.
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: NDamien Lespiau <damien.lespiau@intel.com>
[imre: rebased on patchset version w/o pch/crt/fdi refactoring]
Signed-off-by: NImre Deak <imre.deak@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Just boring sed job for preparation.
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: NDamien Lespiau <damien.lespiau@intel.com>
[imre: rebased on patchset version w/o pch/crt/fdi refactoring]
Signed-off-by: NImre Deak <imre.deak@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Reviewed-by: NDamien Lespiau <damien.lespiau@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Similar to how the ->crtc_mode_set hook should touch the hardware to
enable anything the ->crtc_off hook should disable anything in the
hardware. Otherwise runtime pm for dpms will not work.

Currently the only things left int the haswell_crtc_off hook is
disabling the ddi plls. We can't move the WRPLL enabling out yet
because the current ddi pll sharing code used by the haswell code
doesn't separately track active users and overall users. This must be
fixed by porting it to the generic shared display pll framework, which
is powerful enough.

But the SPLL source is only used by the crt encoder and so can be
moved already. We only need to make sure that the ddi port E is
already off, which hsw_fdi_disable does by calling
intel_ddi_post_disable.

With this the code reorg to shuffle hsw fdi/lpt specific code into a
new hsw-specific crt encoder type is now finally complete.
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
[imre: rebased on patchset version w/o pch/crt/fdi refactoring]
Signed-off-by: NImre Deak <imre.deak@intel.com>
Reviewed-by: NPaulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

The call to intel_ddi_pll_enable in haswell_crtc_mode_set is the only
function that still touches the hardware state from the crtc mode_set
callback on hsw. Since the SPLL isn't ever shared we can easily take
it out into the hsw crt encoder functions.

Temporarily we'll loose a bit of WARN_ON coverage with this, but once
the WRPLLs are switched over that will be restored. For the SPLL
selection add a WARN in the hsw fdi link training code.
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
[imre: rebased on patchset version w/o pch/crt/fdi refactoring]
Signed-off-by: NImre Deak <imre.deak@intel.com>
Reviewed-by: NPaulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

This is needed by an upcoming patch that moves the PCH/CRT PLL disabling
into the post_disable hook, after which we want to keep the modeset
sequence at its current state. At this point this won't have an effect
since the PCH/CRT post_disable hook is atm a NOP.
Signed-off-by: NImre Deak <imre.deak@intel.com>
Reviewed-by: NPaulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

This is needed by an upcoming patch that moves the PCH/CRT PLL enabling
into the pre_enable hook, after which we want to keep the modeset
sequence at its current state. At this point this won't have an effect
since the PCH/CRT pre_enable hook is atm a NOP.
Signed-off-by: NImre Deak <imre.deak@intel.com>
Reviewed-by: NPaulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Luckily the bit definitions match, but it's still confusing
to use one when handling the other. So sprinkle some OCD over
the #defines to make them match and use the right version in
each place.

Maybe we should unify these definitions completely, but that
can always be done sometime in the future.
Reviewed-by: NPaulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

SPLL would be a reference clock we could potentially share,
especially if we want to use the SSC mode. But currently we
don't, so let's rip out this complexity for a simpler conversion
to the new display pll framework.
Reviewed-by: NPaulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

All the other checks also check hw state, so checking our software
refcounts for the plls looks a bit odd. Also this will simplify the
conversion over to the shared dpll framework, which itself has massive
amounts of checks to make sure that we never leave a display pll
enabled when we shouldn't.

So after that conversion we should stil have a good enough coverage of
asserts for entering pc8/runtime pm on hsw/bdw.
Reviewed-by: NPaulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Add !mutex_is_locked() checks to intel_pin_and_fence_fb_obj() and
intel_unpin_fb_obj() to help catch failures to grab struct_mutex when
operating on fb objects.
Signed-off-by: NMatt Roper <matthew.d.roper@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

intel_primary_plane_{setplane,disable} were lacking struct_mutex locking
around their GEM operations.
Signed-off-by: NMatt Roper <matthew.d.roper@intel.com>
Reported-by: NDamien Lespiau <damien.lespiau@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Otherwise we will print some WARNs when we read registers and the
machine is suspended.

Testcase: igt/pm_rpm/debugfs-read
Cc: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: NPaulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

On HSW, the D_COMP register can be accessed through the mailbox (read
and write) or through MMIO on a MCHBAR offset (read only). On BDW, the
access should be done through MMIO on another address. So to account
for all these cases, create hsw_read_dcomp() with the correct
implementation for reading, and also fix hsw_write_dcomp() to do the
correct thing on BDW.

With this patch, we can now get back from the PC8+ state on BDW. We
were previously getting a black screen and lots of dmesg errors.
Please notice that the bug only happens when you actually reach the
PC8+ states, not when you only allow it.

Testcase: igt/pm_rpm/rte
Signed-off-by: NPaulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: NDamien Lespiau <damien.lespiau@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

That function can be used to write anything on D_COMP, not just
disable it, so print a more appropriate message.
Signed-off-by: NPaulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: NDamien Lespiau <damien.lespiau@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

CHV hard hangs on reading on 0x11100

References: https://bugs.freedesktop.org/show_bug.cgi?id=80893Signed-off-by: NMika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>