Due to the necessity of having to take the struct_mutex, the i915
shrinker can not free the inactive lists if we fail to allocate memory
whilst processing a batch buffer, triggering an OOM and an ENOMEM that
is reported back to userspace. In order to fare better under such
circumstances we need to manually retry a failed allocation after
evicting inactive buffers.

To do so involves 3 steps:
1. Marking the backing shm pages as NORETRY.
2. Updating the get_pages() callers to evict something on failure and then
   retry.
3. Revamping the evict something logic to be smarter about the required
   buffer size and prefer to use volatile or clean inactive pages.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

Similar to the madvise() concept, the application may wish to mark some
data as volatile. That is in the event of memory pressure the kernel is
free to discard such buffers safe in the knowledge that the application
can recreate them on demand, and is simply using these as a cache.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

This should help GEM handle memory pressure sitatuions more gracefully.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

There is no need to store the gtt_alignment as it is either explicitly
set according to the hardware requirements (e.g. scanout) or the
minimum alignment is computed on demand.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

If we failed to set the domain, the buffer was no longer being tracked
on any list.

Cc: stable@kernel.org
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

Due to a bogus FBC support check and failing to check for FBC support
in the right places, mode setting on non-mobile platforms could fail
and hang in the FBC disable routine.  Fix it up.

This fix highlights the need for cleanups in this area (function
pointers and better feature support checks).  Patches for that to
follow.
Tested-by: NKenny Graunke <kenny@whitecape.org>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

We now unconditionally restore the mode at lid open time since some
platforms turn off the panel, pipes or other display elements when the
lid is closed.  There's a problem with doing this at resume time
however.

At resume time, we'll get a lid event, but restoring the mode at that
time may not be safe (e.g. if we get the lid event before global state
has been restored), so check the suspended state and make sure our
restore is locked against other mode updates.
Tested-by: NBen Gamari <bgamari.foss@gmail.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

There is a very real possibility that multiple CPUs will notice that the
GPU is wedged. This introduces all sorts of potential race conditions.
Make the wedged flag atomic to mitigate this risk.
Signed-off-by: NBen Gamari <bgamari.foss@gmail.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

This patch uses the previously introduced chip reset logic to reset the
chip when an error event is detected.
Signed-off-by: NBen Gamari <bgamari.foss@gmail.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

This patch puts in place the machinery to attempt to reset the GPU. This
will be used when attempting to recover from a GPU hang.
Signed-off-by: NOwain G. Ainsworth <oga@openbsd.org>
Signed-off-by: NBen Gamari <bgamari.foss@gmail.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

We set a periodic timer to check on the GPU, resetting it every time a
batch is completed. If the timer elapses, we check acthd. If acthd
hasn't changed in two timer periods, we assume the chip is wedged.

This is implemented in such a way that it leaves the option open to
employ adaptive timer intervals in the future. One could wait until
several timer periods have elapsed before declaring the chip dead. If
the chip comes back after several periods but before the "dead"
threshold, the timer interval or dead threshold could be raised.

It is important to note that while checking for active requests, we need
to account for the fact that requests are removed from the list (i.e.
retired) in a deferred work queue handler. This means that merely
checking for an empty request_list is insufficient; the list could be
non-empty yet the GPU still idle, causing the hangcheck timer to
incorrectly mark the GPU as wedged (it took me a while to figure that
out---sigh...)
Signed-off-by: NBen Gamari <bgamari.foss@gmail.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

We'll need it in i915_irq.c for checking whether there are outstanding
requests. Also, the function really ought to return a bool, not an int.
Signed-off-by: NBen Gamari <bgamari.foss@gmail.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

We move the display-specific code into it's own functions, called
from the general GPU state save/restore functions. This will be needed
later by the GPU reset code.
Signed-off-by: NBen Gamari <bgamari.foss@gmail.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

i915_wait_request() only checks mm.wedged after it interacts with the
hardware, generally causing the driver to lock up waiting for a wedged
chip. Make sure we check mm.wedged as the first thing we do.
Reported-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NBen Gamari <bgamari.foss@gmail.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

BLC_PWM_CTL2 is for 965+ only, so add device model check for
legacy backlight control.

For native backlight control, it maps the backlight value (0~255)
in opregion ASLE[BCLP] to backlight duty cycle (0~max_backlight)
and set into control register.

It also add support for IGD device, which follows opregion spec.
Signed-off-by: NLi Peng <peng.li@intel.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

Arrandale has new window based method for panel fitting.
This one enables full screen aspect scaling on LVDS. It fixes
standard mode display failure on LVDS for Arrandale.

Cc: Stable Team <stable@kernel.org>
Signed-off-by: NZhenyu Wang <zhenyuw@linux.intel.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

This is not required on newer stepping hardware to get
reliable force detect status. Removing this fixes screen
blank flicker in CRT detect on IGDNG.

Cc: Stable Team <stable@kernel.org>
Signed-off-by: NZhenyu Wang <zhenyuw@linux.intel.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

IGDNG LVDS SSC uses 120Mhz freq. This fixes one
1600x900 LVDS panel black issue on IGDNG with SSC enabled.

Cc: Stable Team <stable@kernel.org>
Signed-off-by: NZhenyu Wang <zhenyuw@linux.intel.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

New register for PCH LVDS on IGDNG should be used.
This is a copy-n-paste typo. This fixes possible dual
channel LVDS panel failure on IGDNG.

Cc: Stable Team <stable@kernel.org>
Signed-off-by: NZhenyu Wang <zhenyuw@linux.intel.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

We want the compressed line length buffer address, not the framebuffer
address.
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

drm_ht_remove_item() does not handle removing an absent item and the hlist
in particular is incorrectly initialised. The easy remedy is simply skip
calling i915_gem_free_mmap_offset() unless we have actually created the
offset and associated ht entry.

This also fixes the mishandling of a partially constructed offset which
leaves pointers initialized after freeing them along the
i915_gem_create_mmap_offset() error paths.

In particular this should fix the oops found here:
https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/415357/comments/8Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NEric Anholt <eric@anholt.net>
Cc: stable@kernel.org

Ever since we enabled GEM, the pre-9xx chipsets (particularly 865) have had
serious stability issues.  Back in May a wbinvd was added to the DRM to
work around much of the problem.  Some failure remained -- easily visible
by dragging a window around on an X -retro desktop, or by looking at bugzilla.

The chipset flush was on the right track -- hitting the right amount of
memory, and it appears to be the only way to flush on these chipsets, but the
flush page was mapped uncached.  As a result, the writes trying to clear the
writeback cache ended up bypassing the cache, and not flushing anything!  The
wbinvd would flush out other writeback data and often cause the data we wanted
to get flushed, but not always.  By removing the setting of the page to UC
and instead just clflushing the data we write to try to flush it, we get the
desired behavior with no wbinvd.

This exports clflush_cache_range(), which was laying around and happened to
basically match the code I was otherwise going to copy from the DRM.
Signed-off-by: NEric Anholt <eric@anholt.net>
Signed-off-by: NBrice Goglin <Brice.Goglin@ens-lyon.org>
Cc: stable@kernel.org

This patch adds framebuffer compression (good for about ~0.5W power
savings in the best case) support for pre-GM45 chips.  GM45+ have a new,
more flexible FBC scheme that will be added in a separate patch.

FBC can't always be enabled: the compressed buffer must be physically
contiguous and reside in stolen space.  So if you have a large display
and a small amount of stolen memory, you may not be able to take
advantage of FBC.  In some cases, a BIOS setting controls how much
stolen space is available.  Increasing this to 8 or 16M can help.
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: NEric Anholt <eric@anholt.net>

With all the other lid pieces in place, it's easy to generate a uevent
for the LVDS connector just like we do for other outputs.  Should make
lid open/close fit in with the rest of a userland based output
reconfiguration scheme.
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: NEric Anholt <eric@anholt.net>

We can't load or hotplug detect LVDS like we can other outputs, but if
there's a lid device present we can use it as a proxy.  This allows the
LFP state to be determined at ->detect time, making configurations
requiring manual intervention today "just work" assuming the lid device
status is correct.
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: NEric Anholt <eric@anholt.net>

Some laptop platforms will disable pipes and/or planes at lid close time
and not restore them when the lid is opened again.  So catch the lid
event, and if the lid was opened, force a mode restore.

Fixes fdo bug #21230.
Acked-by: NMatthew Garrett <mjg@redhat.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: NEric Anholt <eric@anholt.net>

Some drivers need to know when a lid event occurs and get the current
status.  This can be useful for when a platform firmware clobbers some
hardware state at lid time, and a driver needs to restore things when
the lid is opened again.
Acked-by: NMatthew Garrett <mjg59@srcf.ucam.org>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: NEric Anholt <eric@anholt.net>

Test on the IGD chip, which is a G33-like graphic device.
Signed-off-by: NLi Peng <peng.li@intel.com>
Signed-off-by: NEric Anholt <eric@anholt.net>

When the output device is LVDS, maybe the pixel clock of adjusted_mode will be
less than that in mode. In such case it will set the incorrect multipler factor
in DPLL_MD register.
So the dpll_md_reg will be reset when the output type is non-SDVO

https://bugs.freedesktop.org/show_bug.cgi?id=22761Signed-off-by: NZhao Yakui <yakui.zhao@intel.com>
Reviewd-by: NEric Anholt <eric@anholt.net>
Signed-off-by: NEric Anholt <eric@anholt.net>

Add the missing clone_mask for SDVO-VGA(RGB1)
Signed-off-by: NZhao Yakui <yakui.zhao@intel.com>
Signed-off-by: NEric Anholt <eric@anholt.net>

When the sdvo device is detected as SDVO-LVDS, we will check whether the
brightness is supported by issue SDVO enhancement command.
If it is supported, we will add the brightness property and then brightness
can be adjusted.
Signed-off-by: NZhao Yakui <yakui.zhao@intel.com>
Signed-off-by: NEric Anholt <eric@anholt.net>

When the sdvo device is detected as SDVO-TV, we will check whether the
sepecific picture enhancement is supported. If it is supported, we will
add the corresponnding property for SDVO-TV. We will add the following
property for the SDVO-TV enhancements if they are supported:
 * Contrast/Brightness/Saturation/Hue.
 * left/right/top/bottom margin: This is implemented by using the
   horizontal/vertical overscan enhancements. When the overscan
   enhancements are supported, the above properties will be added. This is
   to be compatible with what we have done in integrated-TV.
 * horizontal pos/vertical pos.

http://bugs.freedesktop.org/show_bug.cgi?id=22891Signed-off-by: NZhao Yakui <yakui.zhao@intel.com>
Signed-off-by: NEric Anholt <eric@anholt.net>

Otherwise, some other userland writing into its buffer may race to land
writes either after the CPU thinks it's got a coherent view, or after its
GTT entries have been redirected to point at the scratch page.  Either
result is unpleasant.
Signed-off-by: NEric Anholt <eric@anholt.net>

Signed-off-by: NFabian Henze <hoacha@quantentunnel.de>
Signed-off-by: NZhenyu Wang <zhenyuw@linux.intel.com>
Signed-off-by: NEric Anholt <eric@anholt.net>

Signed-off-by: NFabian Henze <hoacha@quantentunnel.de>
[Fix reversed HB & IG ids for B43]
Signed-off-by: NZhenyu Wang <zhenyuw@linux.intel.com>
Signed-off-by: NEric Anholt <eric@anholt.net>

Fixes fence timeouts on r6xx/r7xx.  Noticed by
taiu on IRC.
Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>

Got accidently reverted by c93bb85bSigned-off-by: NAlex Deucher <alexdeucher@gmail.com>
Signed-off-by: NDave Airlie <airlied@linux.ie>

This was a non-trivial merge with some patches sent to Linus
in drm-fixes.

Conflicts:
	drivers/gpu/drm/radeon/r300.c
	drivers/gpu/drm/radeon/radeon_asic.h
	drivers/gpu/drm/radeon/rs600.c
	drivers/gpu/drm/radeon/rs690.c
	drivers/gpu/drm/radeon/rv515.c

the pre-r600 fence code returns ebusy if we get hit by a signal
so we should continue to do that.
Signed-off-by: NDave Airlie <airlied@redhat.com>