The IPS driver is designed to be able to run detached from i915 and
just not enable GPU turbo in that case, in order to avoid module
dependencies between the two drivers.  This means that we don't know
what the load order between the two is going to be, and we had
previously only supported IPS after (optionally) i915, but not i915
after IPS.  If the wrong order was chosen, you'd get no GPU turbo, and
something like half the possible graphics performance.
Signed-off-by: NEric Anholt <eric@anholt.net>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: stable@kernel.org

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=25012Reported-by: NTõnu Raitviir <jussuf@linux.ee>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>

It's required by the specs, but we don't know why.  Let's not find out
why.
Signed-off-by: NEric Anholt <eric@anholt.net>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>

Always useful to know just which connector was polled and had its
status updated.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NDave Airlie <airlied@redhat.com>

We were using the lockup struct from the wrong union.
Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Cc: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

On resume, we were attemping to unblank the displays before the
timing and plls had be reprogrammed which led to atom timeouts
waiting for things that are not yet programmed.  Re-program
the mode first, then reset the dpms state.

This fixes the infamous atombios timeouts on resume.
Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Cc: stable@kernel.org
Signed-off-by: NDave Airlie <airlied@redhat.com>

This fixes module reloading and resume as the gfx block seems to
be left in a bad state in some cases.
Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Cc: stable@kernel.org
Signed-off-by: NDave Airlie <airlied@redhat.com>

Only reset the grbm blocks, srbm tends to lock the GPU
if not done properly and in most cases is not necessary.
Also, no need to call asic init after reset the grbm blocks.
Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Cc: stable@kernel.org
Reviewed-by: NJerome Glisse <jglisse@redhat.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

This reverts commit 541cc966.

Wei Yonjun reported this caused a regression against Intel VGA hotplug
on his G33 hw.
Signed-off-by: NDave Airlie <airlied@redhat.com>

Without this, we attempt the handover too late, the firmware fb
might be accessing the chip simultaneously to us re-initializing
various parts of it, which might frighten babies or cause all sort
of nasty psychologic trauma to kitten.
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>
[danvet: add cc: stable, forward ported and compile-fixed for X86]
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
[airlied: move to even earlier in module load.]
Signed-off-by: NDave Airlie <airlied@redhat.com>

This fixes a problem where the wrong card conflicts with vesafb in my
x2 system.
Signed-off-by: NDave Airlie <airlied@redhat.com>

Make sure vram changes hit memory.  This mirrors the
6xx/7xx behavior.
Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

There are too many strange corner cases triggered in old userspace
drivers out there to that it's nearly impossible to not break some
obscure app.

Cc: Jean Delvare <khali@linux-fr.org>
Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Seems to cause problems on certain laptops

Fixes:
https://bugzilla.kernel.org/show_bug.cgi?id=24462Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Tested-by: NLuca Tettamanti <kronos.it@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Fixes the lack of output on the LVDS panel of the Lenovo U160.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=31596Reported-and-tested-by: NDirk Gouders <gouders@et.bocholt.fh-gelsenkirchen.de>
Cc: stable@kernel.org
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>

BIOSes. Can't live without them (apparently), definitely can't live with
them.
Reported-by: NBen Gamari <bgamari@gmail.com>
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=24312Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>

If the tail advances beyond the autoreport HEAD value, then we need to
fallback to an uncached read of the HEAD register in order to ascertain
the correct amount of remaining space in the ringbuffer.
Reported-by: NFang, Xun <xunx.fang@intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=32259Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>

It should be required for all 7xx asics, but seems to cause
problems on some AGP 7xx chips.

Fixes:
https://bugzilla.kernel.org/show_bug.cgi?id=19002Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Reported-and-Tested-by: NDuncan <1i5t5.duncan@cox.net>
Cc: stable@kernel.org
Signed-off-by: NDave Airlie <airlied@redhat.com>

The "e" pointer is either NULL or freed when we call
drm_vblank_put(dev, e->pipe) on the error path.  Just pass the "pipe"
variable directly instead.

I changed another caller to use "pipe" as well for consistency.
Signed-off-by: NDan Carpenter <error27@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Grub doesn't parse spaces in parameters correctly, so
this makes it impossible to force video= parameters
for kms on the grub kernel command line.

v2: shorten the names to make them easier to type.
Reported-by: NSergej Pupykin <ml@sergej.pp.ru>

Cc: Sergej Pupykin <ml@sergej.pp.ru>
Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Cc: stable@kernel.org
Signed-off-by: NDave Airlie <airlied@redhat.com>

The DisplayPort standard (1.1a) states that:
  The I2C-over-AUX Reply field is valid only when Native AUX CH Reply
  field is AUX_ACK (00). When Native AUX CH Reply field is not 00, then,
  I2C-over-AUX Reply field must be 00 and be ignored.

This fixes broken EDID reading when using an active DisplayPort to
duallink DVI converter.  If the AUX CH replier chooses to defer the
transaction, a short read occurs and erroneous data is returned as
the i2c reply due to a lack of length checking and failure to check
for AUX ACK.

As a result, broken EDIDs can look like:
     0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f    0123456789abcdef
00: bc bc bc ff bc bc bc ff bc bc bc ac bc bc bc 45    ???.???.???????E
10: bc bc bc 10 bc bc bc 34 bc bc bc ee bc bc bc 4c    ???????4???????L
20: bc bc bc 50 bc bc bc 00 bc bc bc 40 bc bc bc 00    ???P???.???@???.
30: bc bc bc 01 bc bc bc 01 bc bc bc a0 bc bc bc 40    ???????????????@
40: bc bc bc 00 bc bc bc 00 bc bc bc 00 bc bc bc 55    ???.???.???.???U
50: bc bc bc 35 bc bc bc 31 bc bc bc 20 bc bc bc fc    ???5???1??? ????
60: bc bc bc 4c bc bc bc 34 bc bc bc 46 bc bc bc 00    ???L???4???F???.
70: bc bc bc 38 bc bc bc 11 bc bc bc 20 bc bc bc 20    ???8??????? ???
80: bc bc bc ff bc bc bc ff bc bc bc ff bc bc bc ff    ???.???.???.???.
...

which can lead to:
[drm:drm_edid_block_valid] *ERROR* EDID checksum is invalid, remainder
[drm:drm_edid_block_valid] *ERROR* Raw EDID:
<3>30 30 30 30 30 30 30 32 38 32 30 32 63 63 31 61  000000028202cc1a
<3>28 00 02 8c 00 00 00 00 18 00 00 00 00 00 00 00  (...............
<3>20 4c 61 73 74 20 62 65 61 63 6f 6e 3a 20 33 32   Last beacon: 32
<3>32 30 6d 73 20 61 67 6f 46 00 05 8c 00 00 00 00  20ms agoF.......
<3>36 00 00 00 00 00 00 00 00 0c 57 69 2d 46 69 20  6.........Wi-Fi
<3>52 6f 75 74 65 72 01 08 82 84 8b 96 24 30 48 6c  Router......$0Hl
<3>03 01 01 06 02 00 00 2a 01 00 2f 01 00 32 04 0c  .......*../..2..
<3>12 18 60 dd 09 00 10 18 02 00 00 01 00 00 18 00  ..`.............
Signed-off-by: NDavid Flynn <davidf@rd.bbc.co.uk>
[ickle: fix up some surrounding checkpatch warnings]
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: stable@kernel.org

it has a DSM but the switcher is done via WMI.
Signed-off-by: NDave Airlie <airlied@redhat.com>

Avoid overflowing a 32 bit value.
Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Cc: stable@kernel.org
Signed-off-by: NDave Airlie <airlied@redhat.com>

print the full 64 bit values.
Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Forbid allocating buffer bigger than visible VRAM or GTT, also
properly set lpfn field.

v2 - use max macro
   - silence warning
v3 - don't explicitly set range limit
   - use min macro

Cc: stable <stable@kernel.org>
Signed-off-by: NJerome Glisse <jglisse@redhat.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Prevents code that assumes that the encoder is active when asked to be
disabled from dying a horrible death.
Reported-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Kristian Høgsberg <krh@bitplanet.net>
Acked-by: NJesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: NDave Airlie <airlied@redhat.com>

As we may try to power down the link at various times, it is not
necessarily still coupled with an encoder and so we must be careful not
to depend upon an operation that is only valid when the link is still
attached to a pipe.

Fixes regression in 5bddd17f.
Reported-and-tested-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: stable@kernel.org [after applying 5bddd17f]

In order for bos to retire eventually, a request must be sent down the
ring. This is expected, for example, by occlusion queries for which mesa
will wait upon (whilst running glean) before issuing more batches and so
the normal activity upon the ring is suspended and we need to emit a
request to clear the idle ring.
Reported-by: NJinjin, Wang <jinjin.wang@intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=30380Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>

Linus Torvalds pointed out that our code was unbalanced when powering on
the panel with respect to the power off sequence in that we were failing
to restore the panel-fitter. The consequence of this would be that
across a simple DPMS off/on for a non-native mode, without an intervening
modeset, the panel fitter would remain disabled and the output would shift
on the panel.
Reported-by: NLinus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>

There's not much we can do here but hope for the best. However the first
failure happens quite frequently and if often remedied by the second
attempt to reset HEAD. So only print the error if that attempt also
fails.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=19802Reported-by: NThomas Meyer <thomas@m3y3r.de>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: stable@kernel.org

Otherwise we can't really fix the abi-braindeadness of forcing
libva to manually wait for rendering when switching rings. Which
in turn makes implementing hw semaphores a pointless exercise
(at least for ironlake).

[Also added the relaxed fencing param to explain the jump in
numbering - relaxed fencing is in -next.]
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>

Fixes the modesetting on the secondary panel of the Libretto W100 and
presumably many more Ironlake laptops with SDVO LVDS displays.
Reported-and-tested-by: NMatthew Willoughby <mattfredwill@gmail.com>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: stable@kernel.org

Use the hardware DDA to calculate the ratio with as much accuracy as is
possible.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: stable@kernel.org

If we leave the registers in a conflicting state then when we attempt
to teardown the active mode, we will not disable the pipes and planes
in the correct order -- leaving a plane reading from a disabled pipe and
possibly leading to undefined behaviour.
Reported-and-tested-by: NAndy Whitcroft <apw@canonical.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=32078Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: stable@kernel.org

This workaround only applies to Ironlake.
Signed-off-by: NEric Anholt <eric@anholt.net>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: stable@kernel.org

The pipe is always set to 8BPC, but here we were leaving whatever
previous bits were set by the BIOS in place.
Signed-off-by: NEric Anholt <eric@anholt.net>
Tested-by: NKeith Packard <keithp@keithp.com>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: stable@kernel.org

fixes:
https://bugzilla.kernel.org/show_bug.cgi?id=23752Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Cc:stable@kernel.org
Signed-off-by: NDave Airlie <airlied@redhat.com>

Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Cc: stable@kernel.org
Signed-off-by: NDave Airlie <airlied@redhat.com>

6xx/7xx was hitting the wrong BUS_CNTL reg and bits.
Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Cc: stable@kernel.org
Signed-off-by: NDave Airlie <airlied@redhat.com>