Closes a very unlikely race that can occur if another NonStallInterrupt
method passes between checking fences and acking the previous interrupt.

With this change, the interrupt will re-fire under such conditions.
Tested-by: Tobias Klausmann <tobias.johannes.klausmann@mni.thm.de>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

Commit 1dce6264 introduced a regression
spotted on several G94 (FDObz #85160). This device seems to expect the
vblank period to be set after setting scale instead of before.

V2: shove this in a separate function

This is a candidate bug-fix for 3.18
Signed-off-by: Roy Spliet <rspliet@eclipso.eu>
Tested-by: Zlatko Calusic <zcalusic@bitsync.net>
Tested-by: Michael Riesch <michael@riesch.at>
Tested-by: "poma" <pomidorabelisima@gmail.com>
Tested-by: Adam Williamson <adamw@happyassassin.net>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

Commit "ltc/gf100-: fix cbc issues on certain boards" moved the setting
of the large page size bit from bar/nvc0 to fb/nvc0. GK20A uses its own
FB device and the change was thus not applied to it - fix this.
Signed-off-by: Alexandre Courbot <acourbot@nvidia.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

Extends the fix in f2f9a2cb to also
workaround permission issues noticed by people using AGP systems.

Cc: stable@vger.kernel.org # 3.16: f2f9a2cb: drm/nouveau: fix regression
Cc: stable@vger.kernel.org # 3.16+
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

The binary driver modifies the default context to have this value, rather
than 0x3d0040, *after* it's filled the buffer with the usual golden data.

We don't really have anything in place to locate the correct offset to do
these type of modifications outside of the generation function, so this
will have to do.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

fixups for nouveau and fencing

* 'for-airlied-next' of git://people.freedesktop.org/~mlankhorst/linux:
  drm/nouveau: export reservation_object from dmabuf to ttm
  drm/ttm: add reservation_object as argument to ttm_bo_init
  drm: Pass dma-buf as argument to gem_prime_import_sg_table
  drm/nouveau: assign fence_chan->name correctly
  drm/nouveau: specify if interruptible wait is desired in nouveau_fence_sync
  drm/nouveau: bump driver patchlevel to 1.2.1

Ok, here's the update core-stuff pull request with the locking fixup patch
fixed up with another patch.

* tag 'topic/core-stuff-2014-09-29' of git://anongit.freedesktop.org/drm-intel:
  drm: Drop grab fpriv->fbs_lock in drm_fb_release
  drm/udl: use container_of to resolve udl_fbdev from drm_fb_helper
  drm/ast: use container_of to resolve ast_fbdev from drm_fb_helper
  drm/gma500: use container_of to resolve psb_fbdev from drm_fb_helper
  drm/qxl: use container_of to resolve qxl_fbdev from drm_fb_helper
  drm/nouveau: use container_of to resolve nouveau_plane from drm_plane
  drm/nouveau: use container_of to resolve nouveau_fbdev from drm_fb_helper
  drm/radeon: use container_of to resolve radeon_fbdev from drm_fb_helper
  drm/mgag200: use container_of to resolve mga_fbdev from drm_fb_helper
  drm/cirrus: use container_of to resolve cirrus_fbdev from drm_fb_helper
  drm: Improve debug output for drm_wait_one_vblank
  drm: Fixup locking for universal cursor planes
  drm: Don't update vblank timestamp when the counter didn't change

Adds an extra argument to nouveau_bo_new, which is only used in nouveau_prime.c.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com>

This allows importing reservation objects from dma-bufs.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com>

Allows importing dma_reservation_objects from a dma-buf.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com>

Make nouveau_fence_chan refcounted, to make trace_fence_destroy
always return the correct name without a race condition.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com>

Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com>

Allows userspace to detect shared fences are supported.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com>

This fixes a regression introduced by "drm/nouveau: rework to new fence interface"
(commit 29ba89b2).

The fence sequence should not be reset after creation, the old value is used instead.
On destruction the final value is written, to prevent another source of accidental
wraparound in case of a channel being destroyed after a hang, and unblocking any other
channel that may wait on the about-to-be-deleted channel to signal.

I'm nothing if not optimistic about any hope of recovery from that. ;-)
Reported-by: Ted Percival <ted@tedp.id.au>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com>
Tested-by: Ted Percival <ted@tedp.id.au>
Acked-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

Paulo Zanoni reported a lockdep splat with a locking inversion between
fpriv->fbs_lock and the modeset locks. This issue was introduced in

commit f2b50c11
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date:   Fri Sep 12 17:07:32 2014 +0200

    drm: Fixup locking for universal cursor planes

This here is actually one of the rare cases where lockdep hits a false
positive: The deadlock only happens in drm_fb_release, which cleans up
the file private structure when all the references are gone. So the
locking is the very last one and no one else can deadlock. It also
doesn't protect anything at all, since all ioctls are guaranteed to
have returned at this point - otherwise they'd still hold a reference
on the file.

So let's just drop it and replace it with a big comment.

Cc: David Herrmann <dh.herrmann@gmail.com>
Cc: Matt Roper <matthew.d.roper@intel.com>
Cc: Paulo Zanoni <przanoni@gmail.com>
Reported-and-Tested-by: Paulo Zanoni <przanoni@gmail.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>

Use container_of instead of casting first structure member.
Signed-off-by: Fabian Frederick <fabf@skynet.be>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

Use container_of instead of casting first structure member.
Signed-off-by: Fabian Frederick <fabf@skynet.be>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

Use container_of instead of casting first structure member.
Signed-off-by: Fabian Frederick <fabf@skynet.be>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

Use container_of instead of casting first structure member.
Signed-off-by: Fabian Frederick <fabf@skynet.be>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

Use container_of instead of casting first structure member.
Signed-off-by: Fabian Frederick <fabf@skynet.be>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

Use container_of instead of casting first structure member.
Signed-off-by: Fabian Frederick <fabf@skynet.be>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

Use container_of instead of casting first structure member.
Signed-off-by: Fabian Frederick <fabf@skynet.be>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

Use container_of instead of casting first structure member.
Signed-off-by: Fabian Frederick <fabf@skynet.be>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

Use container_of instead of casting first structure member.
Signed-off-by: Fabian Frederick <fabf@skynet.be>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

This replicates what we've done in i915 in

commit 31e4b89a
Author: Damien Lespiau <damien.lespiau@intel.com>
Date:   Mon Aug 18 13:51:00 2014 +0100

    drm/i915: Print the pipe on which the vblank wait times out

to make sure that when we switch i915 to drm_wait_one_vblank that the
debug output doesn't regress.

Cc: Damien Lespiau <damien.lespiau@intel.com>
Cc: Thomas Wood <thomas.wood@intel.com>
Reviewed-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>

Bunch of things amiss:
- Updating crtc->cursor_x/y was done without any locking. Spotted by
  David Herrmann.
- Dereferencing crtc->cursor->fb was using the wrong lock, should take
  the crtc lock.
- Grabbing _all_ modeset locks torpedoes the reason why we added
  fine-grained locks originally: Cursor updates shouldn't stall on
  background stuff like probing outputs.

Best is to just grab the crtc lock around everything and drop all the
other locking. The only issue is that we can't switch planes between
crtcs with that, so make sure that never happens when someone uses
universal plane helpers. This shouldn't be a possible regression ever
since legacy ioctls also only grabbed the crtc lock, so switching
crtcs was never possible for the underlying plane object. And i915
(the only user of universal cursors thus far) has fixed cursor->crtc
links.

Cc: David Herrmann <dh.herrmann@gmail.com>
Cc: Pallavi G<pallavi.g@intel.com>
Cc: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Tested-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: David Herrmann <dh.herrmann@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>

If we already have a timestamp for the current vblank counter, don't
update it with a new timestmap. Small errors can creep in between two
timestamp queries for the same vblank count, which could be confusing to
userspace when it queries the timestamp for the same vblank sequence
number twice.

This problem gets exposed when the vblank disable timer is not used
(or is set to expire quickly) and thus we can get multiple vblank
disable<->enable transition during the same frame which would all
attempt to update the timestamp with the latest estimate.

Testcase: igt/kms_flip/flip-vs-expired-vblank
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Mario Kleiner <mario.kleiner.de@gmail.com>

v2:Mario: Trivial rebase on top of current drm-next (13-Sep-2014)
Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

Only !P can be used together with a function list.

Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

v2: Don't forget git add, noticed by David.

Cc: David Herrmann <dh.herrmann@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Acked-by: David Herrmann <dh.herrmann@gmail.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

In my header cleanup I've missed the debugfs functions completely.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: David Herrmann <dh.herrmann@gmail.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

Somehow I've missed these three, fix this up asap. Plus move
drm_master_create since while at it.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: David Herrmann <dh.herrmann@gmail.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

Leftover from my previous header cleanup.

This depends upon the patch to rework exynos mmap support, otherwise
it'll break exynos.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Reviewed-by: David Herrmann <dh.herrmann@gmail.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

Now that we've removed the copypasted users in gem/ttm we can
relegate the legacy buffer mapping support to where it belongs.
Also give it the proper drm_legacy_ prefix.

While at it statify drm_mmap_locked, somehow I've missed that in my
previous header rework.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Reviewed-by: David Herrmann <dh.herrmann@gmail.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

The only user I could dig out was i915 back when ums+gem was still a
thing. But we've just very much killed that, and even when someone
screams about that we should resurrect that with a special hack
(wrapping drm_gem_mmap) in i915, not in the core code.

So good riddance to another entry point of the legacy buffer mapping
code.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Reviewed-by: David Herrmann <dh.herrmann@gmail.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

Really, the legacy buffer api should be dead, especially for all these
newfangled drivers. I suspect this is copypasta from the transitioning
days, which probably originated in radeon.

Cc: "Christian König" <christian.koenig@amd.com>
Cc: David Herrmann <dh.herrmann@gmail.com>
Cc: Rashika <rashika.kheria@gmail.com>
Cc: Josh Triplett <josh@joshtriplett.org>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Fabian Frederick <fabf@skynet.be>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Cc: Alexandre Courbot <acourbot@nvidia.com>
Cc: Maarten Lankhorst <maarten.lankhorst@canonical.com>
Cc: Christian Engelmayer <cengelma@gmx.at>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Reviewed-by: David Herrmann <dh.herrmann@gmail.com>
Acked-by: Ben Skeggs <bskeggs@redhat.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

The return value is not used by callers of this function
nor by uses of the DRM_ERROR macro so change the function
to return void.
Signed-off-by: Joe Perches <joe@perches.com>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

Today, most callers of ttm_io_prot() check TTM_PL_FLAG_CACHED before
calling it since on some archs it will unconditionally create non-cached
mappings.

But not all callers do which is incorrect as far as I can tell.

Instead, move that check inside ttm_io_port() itself for all archs
and make powerpc use the same implementation as ia64 and arm
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>

What the code does is equivalent to the x86 code, so let's use
it as well
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>

Merge branch 'exynos-drm-next' of git://git.kernel.org/pub/scm/linux/kernel/git/daeinki/drm-exynos into drm-next

   Sorry for late. This pull request includes some enhancements
   for Exynos drm, new feature supports, cleanups and fixups
   like below,

   - Consider low power transmission for drm mipi dsi module,
     and also add non-continuous clock mode support for Exynos
     mipi dsi driver.
   - Add Exynos3250 SoC support.
   - Enhance and clean up ipp framework and fimc driver.
   - Update to use component match support and fix up
     de-initialization order.
   - Remove a direct mmap interface and relevant stuff specific to
     Exynos drm, use drm generic mmap interface instead.
     And we will remove the specific interface from userspace
     library, libdrm soon.
   - Use universal plane which allows to replace fake primary plane
     with the real one.
   - Some code cleanups and fixups.

* 'exynos-drm-next' of git://git.kernel.org/pub/scm/linux/kernel/git/daeinki/drm-exynos: (40 commits)
  drm/exynos: switch to universal plane API
  drm/exynos: use drm generic mmap interface
  drm/exynos: remove DRM_EXYNOS_GEM_MAP_OFFSET ioctl
  drm/exynos: factor out initial setting of each driver
  drm/exynos/hdmi: unregister connector on removal
  drm/exynos/dp: unregister connector on removal
  drm/exynos/dpi: unregister connector and panel on removal
  drm/exynos/dsi: unregister connector on removal
  drm/exynos/fb: free exynos framebuffer on error
  drm/exynos/fbdev: fix fbdev gem object cleanup
  drm/exynos: fix drm driver de-initialization order
  drm/exynos/ipp: traverse ipp drivers list safely
  drm/exynos: update to use component match support
  drm/exynos/ipp: add file checks for ioctls
  drm/exynos/ipp: remove file argument from node related functions
  drm/exynos/fimc: fix source buffer registers
  drm/exynos/fimc: simplify buffer queuing
  drm/exynos/fimc: do not enable fimc twice
  drm/exynos/fimc: avoid clearing overflow bits
  drm/exynos/ipp: remove events during command cleaning
  ...

The patch replaces legacy functions
drm_plane_init() / drm_crtc_init() with
drm_universal_plane_init() and drm_crtc_init_with_planes().
It allows to replace fake primary plane with the real one.
Additionally the patch leaves cleanup of crtcs to core,
this way planes and crtcs are cleaned in correct order.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>