Remove some inline keywords. One of the functions has clearly outgrown
it anyway, so let's just leave it to the compiler.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

The check for start + length >= total_vm_size is
wrong since start + length can be exactly the size of
the vm.

Fix the check to allow allocation to boundary.

Fixes a regression in commit 4dd738e9
("drm/i915: Fix 32b overflow check in gen8_ppgtt_alloc_page_directories")

Testcase: igt/gem_evict_everything/swapping-interruptible
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90399Tested-by: Lu Hua <huax.lu@intel.com>
Cc: Chris Wilson <chris@chris.wilson.co.uk>
Cc: Dave Gordon <david.s.gordon@intel.com>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Michel Thierry <michel.thierry@intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

During check_crtc_state, scaler_id mispatch is being reported for HSW.
This is applicable for skl+ and not for HSW. It is introduced by
commit id:
    commit a1b2278e
    Author: Chandra Konduru <chandra.konduru@intel.com>
    Date:   Tue Apr 7 15:28:45 2015 -0700

        drm/i915: skylake panel fitting using shared scalers

This patch will make sure that we leave scaler_id as 0 for platforms
before skl and set for skl+ only. This way scaler_id check during
check_crtc_state will pass for both prior to skl and skl+ platforms.

v2:
-Leave scaler_id as 0 for gen < 9 (Daniel)
Signed-off-by: Chandra Konduru <chandra.konduru@intel.com>
References: http://lists.freedesktop.org/archives/intel-gfx/2015-May/065741.htmlSigned-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

As we perform the mmio-flip without any locking and then try to acquire
the struct_mutex prior to dereferencing the request, it is possible for
userspace to queue a new pageflip before the worker can finish clearing
the old state - and then it will clear the new flip request. The result
is that the new flip could be completed before the GPU has finished
rendering.

The bugs stems from removing the seqno checking in
commit 536f5b5e
Author: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Date:   Thu Nov 6 11:03:40 2014 +0200

    drm/i915: Make mmio flip wait for seqno in the work function
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

Synchronising to an object active on the same ring is a no-op, for the
benefit of execbuffer scheduler. However, for CS flips this means that
we can forgo checking whether the last write request of the object is
actually queued and more importantly whether the cache flush for the
write was emitted.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

We no longer interpolate domains in the same manner, and even if we did,
we should trust setting either of the other write domains would trigger
an invalidation rather than force it. Remove the tweaking of the
read_domains since it serves no purpose and use
i915_gem_object_wait_rendering() directly.

Note that this goes back to

commit a8198eea
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date:   Wed Apr 13 22:04:09 2011 +0100

    drm/i915: Introduce i915_gem_object_finish_gpu()

and gpu domain tracking died in

commit cc889e0f
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date:   Wed Jun 13 20:45:19 2012 +0200

    drm/i915: disable flushing_list/gpu_write_list

which is more than 1 year older.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
[danvet: Add notes with information dug out of git history.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

If an batch ends while the IRQs are not turned on the notification can
go missing and the GPU can hang. So generate a warning in this case.
Signed-off-by: Peter Antoine <peter.antoine@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

In skylake update plane functions, intel_tile_height() is called with
bits_per_pixel instead of pixel_format. Correcting it.
Signed-off-by: Chandra Konduru <chandra.konduru@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
[danvet: Fixup alignment.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

After feedback from the hardware team, now we set the GPU min/idel freq to RPe.
Punit is expecting us to operate GPU between Rpe & Rp0. If we drop the
frequency to RPn, punit is failing to change the vgg input voltage to
minimum :(

Since Punit validates the rps range [RPe, RP0]. This patch
removes unused cherryview_rps_min_freq function.

v2: Change commit message

v3: set min_freq before idle_freq (chris)

v4: Squash 'Remove unused rps min function' patch
Signed-off-by: Deepak S <deepak.s@linux.intel.com>
Acked-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

It is observed on BSW that requesting a new frequency from Punit
does nothing when the GPU is in rc6, and if we let GPU enter rc6
with a high frequency, Vnn remains slightly higher than at minimum
frequency. Extending vlv_set_rps_idle() workaround on CHV/BSW.

v2: Update commit msg (Ville)
suggested-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Deepak S <deepak.s@linux.intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

When GPU is idle on VLV, Request freq to punit should be good enough to
get the voltage back to VNN. Also, make sure gfx clock force applies
before requesting the freq fot vlv.

v2: Do forcewake before setting idle frequency (ville)
    Update function comments to match the code (Deepak)

v3: Fix get/put across idle frequency Request. (Ville)

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=75244suggested-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: Deepak S <deepak.s@linux.intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

As we're doing throughout the code, being optimistic that platform n + 1
will mostly reuse the same things as platform n allows us to minimize
the enabling work needed.

This time, it's about the number of WM levels.
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

With the recent modeset internal rework, we wind up setting crtc_state->enable
to false, but leave crtc_state->active as true following a
drmModeSetCrtc(fb=0), which is incorrect.  This mismatch gets caught by
drm_atomic_crtc_check() and causes subsequent atomic operations (such as plane
updates while the CRTC is disabled) to fail.

Bisect points to

        commit dad9a7d6d96630182fb52aae7c3856e9e7285e13
        Author: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
        Date:   Tue Apr 21 17:13:19 2015 +0300

            drm/i915: Use atomic helpers for computing changed flags

as the commit that actually triggers the regression.

The difference compared to (which this patch reverts)

commit 90d46906
Author: Matt Roper <matthew.d.roper@intel.com>
Date:   Thu May 7 14:31:28 2015 -0700

    drm/i915: Set crtc_state->active to false when CRTC is disabled (v2)

is that we know keep state->active/enable in sync for all legacy
modeset paths, as it should be.

Cc: Matt Roper <matthew.d.roper@intel.com>
Cc: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Reported-and-Tested-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
[danvet: Directly squash in the revert and augment the commit
message.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

Revert "drm/i915: Set crtc_state->active to false when CRTC is disabled (v2)"

This reverts commit 90d46906.

Skylake nv12 format requires dbuf (aka. ddb) calculations
and programming for each of y and uv sub-planes. Made minor
changes to reuse current dbuf calculations and programming
for uv plane. i.e., with this change, existing computation
is used for either packed format or uv portion of nv12
depending on incoming format. Added new code for dbuf
computation and programming for y plane.

This patch is a pre-requisite for adding NV12 format support.
Actual nv12 support is coming in later patches.
Signed-off-by: Chandra Konduru <chandra.konduru@intel.com>
Reviewed-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

This patch fix spelling typo in intel_runtime_pm.c
Signed-off-by: Masanari Iida <standby24x7@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

Lost in

commit c5ad54cf
Author: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Date:   Wed May 6 14:36:09 2015 +0300

    drm/i915: Use partial view in mmap fault handler

Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>

We picked up a silent conflict in amdkfd with drm-fixes and drm-next,
backmerge v4.1-rc5 and fix the conflicts
Signed-off-by: Dave Airlie <airlied@redhat.com>

Conflicts:
	drivers/gpu/drm/drm_irq.c

- Add the interrupts & events modules, including new IOCTLs to create and wait
  on events. The HSA RT open source stack is mainly using events to know when
  a dispatched work has been completed. In addition, this module is
  a pre-requisite for the next module I'm going to upstream - debugger support

  This module also handles H/W exceptions, such as memory exception received
  through the IOMMUv2 H/W and Bad Opcode exception receieved from the GPU.

- Adding a new kernel module parameter to let the user decide whether he wants
  to receive a SIGTERM when a memory exception occurs inside the GPU kernel and
  the HSA application doesn't wait on an appropriate event, or if he just want
  to receive notification about this event in dmesg. The default is the latter.

- Additional improvements for SDMA code

- Update my email address in Maintainers file.

* tag 'drm-amdkfd-next-2015-05-19' of git://people.freedesktop.org/~gabbayo/linux:
  drm/amdkfd: change driver version to 0.7.2
  drm/amdkfd: Implement events IOCTLs
  drm/amdkfd: Add module parameter of send_sigterm
  drm/amdkfd: Add bad opcode exception handling
  drm/amdkfd: Add memory exception handling
  drm/amdkfd: Add the events module
  drm/amdkfd: add events IOCTL set definitions
  drm/amdkfd: Add interrupt handling module
  drm/radeon: Add init interrupt kfd->kgd interface
  MAINTAINERS: update amdkfd Oded's email address
  drm/amdkfd: make the sdma vm init to be asic specific
  drm/amdkfd: Use new struct for asic specific ops
  drm/amdkfd: reformat some debug prints
  drm/amdkfd: Remove unessary void pointer cast

Scattering of random drm core patches. Bunch of atomic prep work too, but
the final bits for blob properties, atomic modesets and lifting the
experimental tag on the atomic ioctl are still blocked on Daniel Stone
finalizing and testing the weston support for it. I hope that we can get
it all ready for 4.2 though.

* tag 'topic/drm-misc-2015-05-19' of git://anongit.freedesktop.org/drm-intel: (22 commits)
  drm/atomic: Allow drivers to subclass drm_atomic_state, v3
  drm/atomic: remove duplicated assignment of old_plane_state
  drm/dp: Fix comment in DP helper
  drm/atomic: add drm_atomic_get_existing_*_state helpers
  drm/core: get rid of -Iinclude/drm
  drm/i915: get rid of -Iinclude/drm
  drm/atomic-helpers: Export drm_atomic_helper_update_legacy_modeset_state
  drm/atomic-helpers: Update vblank timestamping constants
  drm/sysfs: remove unnecessary connector type checks
  drm/sysfs: split DVI-I and TV-out attributes
  drm/sysfs: make optional attribute groups per connector type
  drm/sysfs: add a helper for extracting connector type from kobject
  drm/edid: Add CEA modes before inferred modes
  drm/prime: Allow internal imports without import_sg_table
  drm: Add reference counting to blob properties
  drm: Introduce blob_lock
  drm: Introduce helper for replacing blob properties
  drm: Don't leak path blob property when updating
  drm/atomic: Don't open-code CRTC state destroy
  drm/edid: Add DMT modes with ID > 0x50
  ...

Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>

Signed-off-by: Andrew Lewycky <Andrew.Lewycky@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>

This patch adds a new kernel module parameter to amdkfd,
called send_sigterm.

This parameter specifies whether amdkfd should send the
SIGTERM signal to an HSA process, when the following conditions
occur:

1. The GPU triggers an exception regarding a kernel that was
   issued by this process.

2. The HSA process isn't waiting on an event that handles
   this exception.

The default behavior is not to send a SIGTERM and suffice
with a dmesg error print.
Reviewed-by: Ben Goz <ben.goz@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>

Signed-off-by: Alexey Skidanov <alexey.skidanov@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>

This patch adds Peripheral Page Request (PPR) failure processing
and reporting.

Bad address or pointer to a system memory block with inappropriate
read/write permission cause such PPR failure during a user queue
processing. PPR request handling is done by IOMMU driver notifying
AMDKFD module on PPR failure.

The process triggering a PPR failure will be notified by
appropriate event or SIGTERM signal will be sent to it.

v3:
- Change all bool fields in struct kfd_memory_exception_failure to
  uint32_t
Signed-off-by: Alexey Skidanov <alexey.skidanov@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>

This patch adds the events module (kfd_events.c) and the interrupt
handle module for Kaveri (cik_event_interrupt.c).

The patch updates the interrupt_is_wanted(), so that it now calls the
interrupt isr function specific for the device that received the
interrupt. That function(implemented in cik_event_interrupt.c)
returns whether this interrupt is of interest to us or not.

The patch also updates the interrupt_wq(), so that it now calls the
device's specific wq function, which checks the interrupt source
and tries to signal relevant events.

v2:

Increase limit of signal events to 4096 per process
Remove bitfields from struct cik_ih_ring_entry
Rename radeon_kfd_event_mmap to kfd_event_mmap
Add debug prints to allocate_free_slot and allocate_signal_page
Make allocate_event_notification_slot return a correct value
Add warning prints to create_signal_event
Remove error print from IOCTL path
Reformatted debug prints in kfd_event_mmap
Map correct size (as received from mmap) in kfd_event_mmap

v3:

Reduce limit of signal events back to 256 per process
Fix allocation of kernel memory for signal events
Signed-off-by: Andrew Lewycky <Andrew.Lewycky@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>

- AMDKFD_IOC_CREATE_EVENT:
	Creates a new event of a specified type

- AMDKFD_IOC_DESTROY_EVENT:
	Destroys an existing event

- AMDKFD_IOC_SET_EVENT:
	Signal an existing event

- AMDKFD_IOC_RESET_EVENT:
	Reset an existing event

- AMDKFD_IOC_WAIT_EVENTS:
	Wait on event(s) until they are signaled

v2:

- Move the limit of the signal events to kfd_ioctl.h so it
  can be used by userspace

v3:
- Change all bool fields in struct kfd_memory_exception_failure
to uint32_t
Signed-off-by: Andrew Lewycky <Andrew.Lewycky@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>

This patch adds the interrupt handling module, kfd_interrupt.c, and its
related members in different data structures to the amdkfd driver.

The amdkfd interrupt module maintains an internal interrupt ring
per amdkfd device. The internal interrupt ring contains interrupts
that needs further handling. The extra handling is deferred to
a later time through a workqueue.

There's no acknowledgment for the interrupts we use. The hardware
simply queues a new interrupt each time without waiting.

The fixed-size internal queue means that it's possible for us to lose
interrupts because we have no back-pressure to the hardware.

However, only interrupts that are "wanted" by amdkfd, are copied into
the amdkfd s/w interrupt ring, in order to minimize the chances
for overflow of the ring.
Signed-off-by: Andrew Lewycky <Andrew.Lewycky@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>

This patch adds a new interface function to the kfd->kgd interface.
The function is kgd_init_interrupts() and its function is to
initialize a pipe's interrupts.

The function currently enables the timestamp interrupt and the
bad opcode interrupt.
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>

Leaving AMD soon so need to update my email address to @gmail.com
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>

Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>

This patch creates a new structure for asic specific operations, instead
of using the existing structure of operations.

This is done to make the code flow more logic, readable and maintainable.

The change is done only to the device queue manager module at this point.
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>

Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>

kmalloc() returns a void pointer - no need to cast it in
drivers/gpu/drm/amd/amdkfd/kfd_process.c::kfd_process_destroy_delayed()
Signed-off-by: Firo Yang <firogm@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@amd.com>

- skl plane scaler support (Chandra Kondru)
- enable hsw cmd parser (Daniel and fix from Rebecca Palmer)
- skl dc5/6 support (low power display modes) from Suketu&Sunil
- dp compliance testing patches (Todd Previte)
- dp link training optimization (Mika Kahola)
- fixes to make skl resume work (Damien)
- rework modeset code to fully use atomic state objects (Ander&Maarten)
- pile of bxt w/a patchs from Nick Hoath
- (linear) partial gtt mmap support (Joonas Lahtinen)

* tag 'drm-intel-next-2015-05-08' of git://anongit.freedesktop.org/drm-intel: (103 commits)
  drm/i915: Update DRIVER_DATE to 20150508
  drm/i915: Only wait for required lanes in vlv_wait_port_ready()
  drm/i915: Fix possible security hole in command parsing
  drm/edid: Kerneldoc for newly added edid_corrupt
  drm/i915: Reject huge tiled objects
  Revert "drm/i915: Hack to tie both common lanes together on chv"
  drm/i915: Work around DISPLAY_PHY_CONTROL register corruption on CHV
  drm/i915: Implement chv display PHY lane stagger setup
  drm/i915/vlv: remove wait for previous GFX clk disable request
  drm/i915: Set crtc_state->active to false when CRTC is disabled (v2)
  drm/i915/skl: Re-indent part of skl_ddi_calculate_wrpll()
  drm/i915: Use partial view in mmap fault handler
  drm/i915: Add a partial GGTT view type
  drm/i915: Consider object pinned if any VMA is pinned
  drm/i915: Do not make assumptions on GGTT VMA sizes
  drm/i915/bxt: Mark WaCcsTlbPrefetchDisable as for Broxton also.
  drm/i915/bxt: Mark WaDisablePartialResolveInVc as for Broxton also.
  drm/i915/bxt: Mark Wa4x4STCOptimizationDisable as for Broxton also.
  drm/i915/bxt: Move WaForceEnableNonCoherent to Skylake only
  drm/i915/bxt: Enable WaEnableYV12BugFixInHalfSliceChicken7 for Broxton
  ...

Two watchdog changes that came through different trees had a non
conflicting conflict, that is, one changed the semantics of a variable
but no actual code conflict happened. So the merge appeared fine, but
the resulting code did not behave as expected.

Commit 195daf66 ("watchdog: enable the new user interface of the
watchdog mechanism") changes the semantics of watchdog_user_enabled,
which thereafter is only used by the functions introduced by
b3738d29 ("watchdog: Add watchdog enable/disable all functions").

There further appears to be a distinct lack of serialization between
setting and using watchdog_enabled, so perhaps we should wrap the
{en,dis}able_all() things in watchdog_proc_mutex.

This patch fixes a s2r failure reported by Michal; which I cannot
readily explain. But this does make the code internally consistent
again.
Reported-and-tested-by: Michal Hocko <mhocko@suse.cz>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Pull MTD fixes from Brian Norris:
 "Two MTD fixes for 4.1:

   - readtest: the signal-handling code was clobbering the error codes
     we should be handling/reporting in this test, rendering it useless.
     Noticed by Coverity.

   - the common SPI NOR flash DT binding (merged for 4.1-rc1) is being
     revised, so let's change that before 4.1 is minted"

* tag 'for-linus-20150516' of git://git.infradead.org/linux-mtd:
  Documentation: dt: mtd: replace "nor-jedec" binding with "jedec, spi-nor"
  mtd: readtest: don't clobber error reports

Drivers may need to store the state of shared resources, such as PLLs
or FIFO space, into the atomic state. Allow this by making it possible
to subclass drm_atomic_state.

Changes since v1:
- Change member names for functions to atomic_state_(alloc,clear)
- Change __drm_atomic_state_new to drm_atomic_state_init
- Allow free function to be overridden too, in case extra memory is
  allocated in alloc.
Changes since v2:
- Rename *_default_free to default_release, to make clear it doesn't
  free the state object itself.

Cc: dri-devel@lists.freedesktop.org
Acked-by: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

old_plane_state is already assigned to old_state->plane_states[i] inside
for_each_plane_in_state(). Here we remove an the extra assignment.
Signed-off-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>